fib flow
USAOACS Technical Library
5 0712 01014934 1
Research and Development Technical Report
ECOM- 74-01 04-F
technical
LIBRARY
EVALUATION OF Nd:YVO, AND Ho:Er:Tm: YVO, AS
4 4
PULSE PUMPED Q-SWITCHED LASERS
M. Bass
L. G. DeShazer
P. P. Yaney
Center for Laser Studies
University of Southern California
Los Angeles, California 90007
I January 1976
• Final Report for Period 1 January 1974 - 1 April, 1975
• Distribution Statement
•
t Approved for public release;
• distribution unlimited.
Prepared for
ECOM
US ARMY ELECTRONICS COMMAND FORT MONMOUTH
NEW JERSEY 07703
CJC-FM J7 00-45
NOTICES
Disclaimers
The findings in this report are not to bo construed as an
official Department of the Army position, unless so desig-
nated by other authorized documents.
The citation of trade names and names of manufacturers in
this report is not to be construed as official Government
indorsement or approval of commercial products or services
referenced herein.
Disposition
Destroy this report when it is no longer needed. Do not
return it to the originator.
L
UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PACE (Wt ,an Dat a Enlarad)
REPORT DOCUMENTATION PAGE
READ INSTRUCTIONS
BEFORE COMPLETING FORM
1 REPORT NUMBER
ECOM-74-OI 04-F
2. GOVT ACCESSION NO.
3. RECIPIENT'S CATALOG NUMBER
« TITLE (and Subtitle)
EVALUATION OF Nd:YV04 AND
Ho:Er:Tm:YVOj AS PULSE PUMPED
S. TYPE OF REPORT ft PERIOD COVERED
Final Report
1 Jan 74-1 Aor 75
Q-SWITCHED LASERS
6. PERFORMING ORG. REPORT NUMBER
7. AUTHORf*;
Michael Bass, Larry G. DeShazer, and
Perry P. Yaney
8. CONTRACT OR GRANT NUMBER^;
DA A B 07- 74- C- 01 04
9 PERFORMING ORGANI ZATION NAME AND ADDRESS
Center for Laser Studies
University of Southern California
Los Angeles, CA 90007
10. PROGRAM ELEMENT, PROJECT. TASK
AREA ft WORK UNIT NUMBERS
1S7 62703 A 1 86 06
11. CONTROLLING OFFICE NAME AND ADDRESS
US Army Electronics Command
12. REPORT DATE
January 1 976
ATTN: AMSEL-CT-L-D
13. NUMBER OF PAGES
U MONITORING AGENCY NAME ft ADDRESS^// dlttarani from Controlling O ftlca)
15. SECURITY CLASS, (ot thla raporl)
Unclas sified
15a. DECLASSI FI CATION/ DOWN GRADING
SCHEDULE
16 DISTRIBUTION STATEMENT (ot thla Report)
Approved for Public Release; Distribution of this
i document is Unlimited.
17. DISTRIBUTION STATEMENT (ot tha abatracl anlarad In Block 20, If dlftoront trom Raport)
18 supplementary notes
None
19 KEY WORDS (Contln ua on ravaraa alda it nacaaaary and Identify by block numbar)
Yttrium orthovanadate (YVO^); Lasers; Neodymium (Nd); Holmium (Ho),
Erbium (Er), Thulium (Tm); Absorption, Emission Spectra; Energy
Levels, Stimulated Emission Cross Sections; Inclusions, Irridium;
Crystal Growth
20 ABSTRACT (Contlnua on ravaraa alda It nacaaaary and Idantlty by block numbar)
This program was designed to develop and exploit rare-earth doped
yttrium orthovanadate (YVO^) as a high efficiency laser material. The
effort included crystal growth, spectroscopy and laser performance testing*
Though emphasis was placed on the spectroscopic and laser properties of
Nd^+ ions, the spectroscopy of Hc?+ ions with Er^ and Tm^+ co-dopants
was also studied* Measurements of the optical absorption and emission
spectra, enerev levels and stimulated emission cross-sections are renorted
DD , 1 473 EDITION OF 1 NOV 65 IS OBSOLETE
S/N 0 102-014- 6601 |
SECURITY CLASSIFICATION OF THIS PAGE (Whan Da fa Kntarad)
Unclassified
-LLUMITY CLASSIFICATION OF THIS P AGEfWhen Dili Enttrmd)
20. (Continued)
Details in the spectroscopy of Nd^+ in YVO4 were measured and analyzed.
The Cary 17 Spectrophotometer was used to measure the absorption
spectrum of Nd^+:YV04 from 0. 38 to 2. 0 Mm. The most intense absorption
line in the 0.81 |jm absorption group, the group most appropriate for diode
pumping, was the line at 0. 8080 |im with a = 9 cm ^ . The F 3 / 2 ”♦ ^Tg / 2
spectrum is still somewhat of a puzzle. There seems to be more than one
set of spectra for Nd^* in YVO^ suggestive of either a second site for the
Nd^+ ions or else that there is significant Nd-Nd pair spectra present.
Laser emission in Ho doped materials was analyzed theoretically.
The oscillator strengths, transition rates and branching ratios associated
with known and potential laser transitions in Ho^+ were calculated using a
computer.
Since no NdrYVO^ laser rods suitable for testing were delivered by
the subcontractor during the second part of this program no new laser
testing was conducted. Several Ho, Er, TmrYVO^ laser rods were delivered
but were not tested because our emphasis was concentrated on the Nd
problem. Nevertheless, in the first six months of this program it was
demonstrated that NdiYVO^ could outperform Nd:YAG in pulse pumped
Q -switched operation. In addition, reliable, pre-pulse free, single pulse
Q switching of NdrYVO^ was demonstrated with no intracavity polarizer.
SECURITY CLASSIFICATION OF THIS PAGEfHTi.n Dmlm Enttfd)
SUMMARY
Yttrium orthovanadate (YVO^) doped with rare earth ions was investi-
gated for use as a pulse pumped Q switched laser material. The spectros-
3+ 3 +
copy of two potential laser ions, Nd and Ho , was studied in detail.
Measurements of the optical absorption and emission spectra, energy levels
and stimulated emission cross-sections are reported. Q switched lasing of
pulse pumped Nd:YVO^ was demonstrated and found to be superior in output
to that of Nd:YAG in the same pump cavity. The pump cavity was a small
gold plated single ellipse and was designed for optimum Nd:YAG performance.
In addition, the strong inherent birefringence of Nd:YVO^ made possible pre-
pulse free, single pulse Q switching with no- intracavity polarizer.
The crystal growth and materials evaluation effort of this program
demonstrated that large boules (1.3 x 7 cm) of material grown along the crystalo-
graphic A axis could be prepared without visible color centers. It was found
however that inclusions of irridium from the crucible were often present and
were the cause of significant losses in several crystals. These inclusions are
often present along the a-plane which is a cleavage plane of YVO^. Their
presence weakens the bonding and cleavage frequently occurs during the cool-
down. This is a major and as yet unresolved problem. If cool-down was
achieved without severe cleavage, the stresses of sawing, grinding and polish-
ing often caused material failure during rod fabrication.
It is the conclusion of this project that when the growth and fabrication
problems of Nd:YVO^ are solved, this material will be a suitable replacement
for Nd:YAG in laser systems where input power is limited.
iii
PREFACE
This is the final report on the ’’Evaluation of NdrYVO^ and Ho, Er,
TmiYVO^ as Pulse Pumped Q-Switched Lasers”. The research was per-
formed under U. S. Army Electronics Command Contract No. DAAB07-74-
C-0104. The program monitor was Mr. John W. Strozyk, Fort Monmouth,
New Jersey. The research covers the period 2 January 1974 to 30 March
1975 with the results of the work performed during the period from
2 January 1974 to 30 June 1 974, reported in the first semiannual report of
this program (Report # ECOM-74-Ol 04-1 ) and the work performed during
the period from 1 July 1974 to 30 March 1975 described herein.
The principal investigators during this project were Michael Bass,
Associate Director of the Center for Laser Studies, Uri Ranon, Senior
Research Scientist at the Center, and, following Dr. Ranon’ s leaving the
Center, Dr. Larry G. DeShazer, Director of the Center, was substituted
as co-principal investigator.
The following personnel participated in various phases of the study:
Laser Testing
Spectroscopy, Nd
Spectroscopy, Ho
Spectroscopic Analysis
Color Centers
M. Bass, U. Ranon
L. G. DeShazer, M. Bass, J. K. Guha,
P. P. Yaney, K. M. Leung
E. D. Reed and U. Ranon
L. G. DeShazer, P. P. Yaney, J. A. Caird
J.K. Guha
Mr. L. R.Rothrock, R.E. Wilder and Dr. D. Brandle of the Crystal
Products Department, Union Carbide Corporation, San Diego, California,
supplied the spectroscopic samples and laser rods used in this program.
They prepared the portion of this report concerning crystal growth and
laser rod fabrication. Dr. A. B. Chase of the Aerospace Corporation, El
Segundo, California, conducted an evaluation of material quality and the
results of that work are also described in this report.
iv
During this report period, a paper describing some of the work
sponsored by this contract was published. This paper was by J. A. Caird
3-l-
and L. G. DeShazer, '’Analysis of Laser Emission in Ho -Doped Materials','
IEEE J. of Quantum Electronics QE- 1 1 , 97 (1975). Also, several meet-
ings were held to discuss the status of the work:
8 Oct. 1974 M. Bass. L. R Rothrock, and D. Brandle visited J. Strozyk
and E. Schiel at Fort Monmouth, NJ.
16 Jan. 1975
30 Jan. 1975
14 Feb. 1975
L. G. DeShazer visited A.B. Chase at Aerospace Corporation.
L. G. DeShazer visited J. Strozyk and E. Schiel at Fort
Monmouth, NJ.
L. G. DeShazer visited L. R Rothrock at UCC.
v
TABLE OF CONTENTS
REPORT DOCUMENTATION PAGE DD FORM 1473
SUMMARY
PREFACE
TABLE OF CONTENTS
LIST OF ILLUSTRATIONS
LIST OF TABLES
I. INTRODUCTION
II. MATERIALS GROWTH AND PREPARATION
A. Growth of YVO^ Crystals at Union Carbide Corp.
1. Background
2. Melt Effects
3. Growth
4. Fabrication
5. Analysis
III.
B. Evaluation of Scattering Centers in Czochralski Grown
YVO,
4
C. Flux Growth of YVO^ at Aerospace Corporation
SPECTROSCOPY OF RARE EARTH
A. Introduction
B. Experiment
C. Spectra
1.
2.
3.
4I -2P
9/2 1/2
4i 4f
9/2 3/2
4 4
F I
3/2 n/2
D. Analysis
1. Site Symmetry
2. Electric Dipole Transitions Between Crystal-Field
Levels
Page
i
iii
iv
vi
viii
xii
1
4
4
4
4
6
12
15
27
32
37
37
38
41
46
46
48
49
49
51
vi
Page
3. The P State in D_ Symmetry 58
1 / Z Zd
E. Energy Level Scheme 61
F. Summary 68
IV. SURVEY OF Nd: YVO . ABSORPTION 69
« 4
V. LASER PERFORMANCE TESTING 90
VI. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE
EFFORTS 98
REFERENCES 99
3 +
APPENDIX A: Analysis of Laser Emission in Ho -Doped
Materials
APPENDIX B: Continuous -wave Operation of Nd:YVO at 1.06 and
1. 34 M
vii
LIST OF ILLUSTRATIONS
r
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Phase diagram-system After E. M. Levin,
J. Am. Cer. Soc. 50:7, 381 (1967)
Remnants of YVO, Melts
4
Laue patterns for (100) and (110) YVO^
Cross-section habits of C-axis and A -axis YVO.
4
Some boules grown during contract
Included particle (A) and general crystal matrix in
( Y. ,„Er Tm 0Hoft n,)VO,
' 0.17 0.75 0. 07 0. 01 4
E. D. X. A . of area "A" in Figure 6
E. D. X. A. of area "B" in Figure 6
Included particles along a cleavage step in
(Y . ,„Er Tm H0 )VO,
' 0.17 0. 75 0.07 °o. 01 4
E. D. X.A. of area "C" in Figure 9
E. D. X. A . of area " D" in Figure 9
A large inclusion in (YQ. , ?Er0_ ^Trry 07Ho0. 01>VC>4
E. D. X. A . of area "E" in Figure 12
E. D. X. A . of pure powder,
E. D. X. A. of pure E^O^ powder
E. D. X. A. of pure YVO^ powder
Scattering centers in Nd:YVO^
Scattering centers iri Nd:YVO^. The polarization of the illumi-
nating light was parallel to one of the polarization directions of
(100) plane in (a) and to the other in plane (b)
Triangular scattering center in NdrYVO^. Possibly Ir or Pt ..
Normal plate habit of YVO^. Also has small ~normal zircon
habit
Flux grown YVO^ Crystals
Experimental arrangements for absorption spectroscopy
Page
5
7
9
10
11
16
17
18
19
20
21
22
23
24
25
26
28
29
31
34
35
40
viii
LIST OF ILLUSTRATIONS (continued)
Page
23. Polarized absorption spectra of the ^9/2^ ^1/2 trans*ltlons
of Nd-.YVO^ at ~300°K. The letters a, b# c, and e, denote the
crystal field levels of the ^9/2 state# The c-axis was trans-
verse to the observation axis 42
24. Polarized absorption spectra of the ^9/2 “*^3/2 trans*ltlons
NdrYVO^ at two temperatures. The peak of the Flat-topped rr-line
at~ 300°Kwas determined to be at 18.5 cm“', Crystal oriented
as In Figure 23 43
25. The dye-laser -excited, polarized fluorescence spectra of the
^^3/2 "*^9/2 transitions of Nd:YVO^ at two temperatures. The
gain factors xl , etc. f apply only within the given temperature.
The letters a1 and b1 denote the crystal-field levels of the ^Fg/2
state while the unprimed letters identify the levels belonging to the
^Ig/2 state. The inset (i) shows the self-absorption which occurs
when the dye laser beam is moved back from the observed surface.
Crystal oriented as in Figure 23 44
26. The dye-laser-excited, polarized fluorescence spectra of the
4^3/2 "*^*11/2 *rans iti°ns Nd: YVO^ at two temperatures. The
gain factors xl, etc., apply only within the given temperature.
The letters a1 and b1 denote the crystal-field levels of the ^Fg/2
state while the unprimed letters identify the levels belonging to the
^11/2 state* Crystal oriented as in Figure 23 45
27. The tetragonal unit cell Qf yttrium vanadate after Wyckoff
(Ref. 15). The cell contains four YVO4 molecules. Eight face-
oxygen ions associated with four vanadium-oxygen tetrahedra that
are located at the four off-center face positions of vanadium have
been omitted from the cell to enhance clarity 50
28. Energy levels ( + 1 cm~^) of selected states of NdrYVO^ at room
temperature showing transitions resolved at ~85°K. The approxi-
mate positions of the levels at ~85°K are shown as short dashed
lines, (a) This is the ~85°K value. See Table 6 63
29. 2. 5 pm band of Nd:YVO^ absorption spectra at room temperature.
Taken with Beckman DK-2A spectrophotometer. 1% Nd. 4 mm
thick 74
30. 1. 6 pm band of NdrYVO^ absorption spectra at room temperature.
Partially polarized light propagating along the a-axis. Mostly n
spectrum. 1% Nd. 4 mm thick 75
ix
LIST OF ILLUSTRATIONS (continued) Page
31. 1. 6 pm band of Nd:YVO^ absorption spectra at room tempera-
ture. Lower curve is a spectrum. Propagation along c-axis.
1% Nd. 5 mm thick. Upper curves demonstrate the fact that
the Cary 17 spectrophotometer is partially linearly polarized.
Propagation along a-axis. 1% Nd. 4 mm thick 76
32. 0. 89 pm band of NdrYVO^ absorption spectra at room tempera-
ture. a spectrum. 1% Nd. 4 mm thick 77
33. 0. 89 pm band of NdrYVO^ absorption spectra at room tempera-
ture. a spectrum. 3% Nd. 8. 5 mm thick. ••••••••••••••••••• 78
34. 0. 89 pm band of NdrYVO^ absorption spectra at room tempera-
ture. rr spectrum. 1% Nd. 4 mm thick. ••••••• •••••••• 79
35. 0. 89 pm band of NdiYVO^ absorption spectra at room tempera-
ture. TT spectrum. 3% Nd. 8. 5 mm thick 80
36. 0. 75 and 0. 81 pm bands of NdrYVO^ absorption spectra at room
temperature, o spectrum. 1 % Nd. 4 mm thick ............. . 81
37. 0. 75 and 0. 81 pm bands of Nd:YVO^ absorption spectra at room
temperature, tt spectrum. 1% Nd. 4 mm thick 82
38. 0. 59 pm (yellow) band of NdiYVC^ absorption spectra at room
temperature, a spectrum. 1% Nd. 4 mm thick 83
39. 0. 59 pm (yellow) band of NdrYVO^ absorption spectra at room
temperature, n spectrum. 1% Nd. 4 mm thick 84
40. Yellow band of NdrYVO^ absorption spectra at room temperature.
rr spectrum on scale. 1% Nd. 4 mm thick 85
41. 0. 53 pm band of Nd:YVO^ absorption spectra at room tempera-
ture. a spectrum. 1% Nd. 4 mm thick 86
42. 0. 53 pm band of Nd:YVO^ absorption spectra at room tempera-
ture. rr spectrum. 1% Nd. 4 mm thick 87
43. X< 0. 5 pm bands of Nd: YVO^ absorption spectra at room tem-
perature. a spectrum. 1% Nd. 4 mm thick 88
44. \< 0. 5 pm bands of Nd: YVO^ absorption spectra at room tem-
perature. tt spectrum. 1% Nd. 4 mm thick 89
45. Nd:YAG laser output vs input energy data with and without the
ignitron in the discharge circuit. Current pulse FWHM is
125 ps with a krypton-filled lamp, 1200 Torr 91
x
LIST OF ILLUSTRATIONS (continued) Page
46. Output versus input energies for the 3 x 30 mm Nd:YAG
(ECOM-YAG) using the Xe flashlamp (450 T) and the 125 jjsec
(FWHM) pump pulse duration. Output R = 45% 92
47. Q-switched laser output vs input energy from Nd:YVO^ (rod 3L)
using a Pockels cell Q-switch and no intracavity polarizer. The
output reflector was 3% R at 1.06 jjm. x Pockels cell Q-switched.
0 Long pulse lasing with Pockels cell in cavity. C3 Long pulse
lasing, empty cavity 93
48. Electrooptically Q-switched configurations, (a) A Pockels -cell-
Glan-polarizer combination in a Nd:YAG laser cavity, (b) A
Pockels cell only in a cavity which has a strong polarization-
dependent gain, (c) A Pockels cell in a cavity where the laser
rod is slightly wedged and strongly birefr ingent. • 94
xi
LIST OF TABLES
1.
2.
3.
4.
Comparison of Neodymium Laser Hosts
Starting Materials Available for Crystal Growth
List of Boules Grown
Transitions Permitted by the 3-j Symbol in Eq. (3) for the
4 4 3 +
L . F„ Transitions of Nd in 2D„ , Symmetry . . . .
9/2 3/2 2d 7 7
5. Electric Dipole Selection Rules in D_ Symmetry
Zd
o 4-
6. Observed Transitions of ~1 Atomic % Nd in YVO . • . . . .
4 4 4
7. Parameters for F^^ "* I \\/2 ^rans ’ltlons ’in Nd:YVO^ at
Room Temperature
8. Absorption Lines of Nd:YVO^ at Room Temperature
Nominally 1% Nd
Page
3
8
13
58
60
64
65
69
xii
I. INTRODUCTION
The object of this program is to investigate rare earth doped yttrium
orthovanadate (YVO^) for use as a low threshold, pulse pumped, Q-switched
laser material. Detailed studies of Nd:YVO„ laser material were made.
4
This included spectroscopic measurements of interest for predicting and
evaluating laser performance and demonstration of long pulse and pulse
pumped, Q-switched operation. Because of the concentration on Nd:YVO^
only spectroscopic measurements were made for the Ho, Er, Tm:YVO^ laser
material.
YVO^ was proposed as a laser host crystal many years ago [ 1, Z]
but was dropped because of early crystal growth difficulties. Recent advances
in the growth of this crystal [3] however show that these difficulties are not
insurmountable. As part of this program large A axis boules (1. 3x7 cm)
and 3 x 30 mm Nd:YVO^ laser rods have been prepared. These have been
free of color centers and have had excellent optical quality. A laser rod
fabrication problem was encountered causing the loss of several potential
laser rods. However, with improved material (i. e. , material free of
included irridium) the fabrication problem should be correctable.
Table 1 summarizes the pertinent data for Nd in YVO^. Included in
Table 1 are data for Nd in YAG, GGG and YAIO^. The peak cross section
for the laser transition at 1 . 06 (am of Nd in YVO^ is 30 x 10“ cm as com-
pared to 6. 5 x 10“19 cm^ in YAG. The fluorescent lifetime for the upper
4
laser level ( in YVO^ is accordingly smaller than in YAG, 92 (a sec as
compared to 240 pisec. In a small elliptical laser cavity, the threshold for
long pulse operation of Nd:YVO^ is less than half that for Nd:YAG. The table
also includes the crystal growth rate (~ 3x higher for Nd:YVO^ than Nd:YAG)
and segregation coefficient (~2X larger in Nd:YVO^ than in Nd:YAG).
Specifically, it is to be noted that this report summarizes a 15 month
study of the laser properties and spectroscopy of Nd:YVO^ and of the spectros-
copy of Ho, Er, TmtYVO^. The materials preparation and quality evaluation
1
efforts are described in Section II. Spectroscopic measurements and their
interpretation are discussed in Section III. Section IV presents a survey
of the NdrYVO^ absorptions. The laser testing, while described in detail
in the semiannual report, is discussed and Pockels cell Q switching of
NdrYVO^ with no intracavity polarizer is analyzed in Section V. In Section
VI, the major conclusions of this program are summarized and recommenda-
tions for further work are given. In Appendix A we present a discussion of
34-
laser emission in Ho doped materials. In addition, for completeness we
have included in Appendix B a description of work done at The Aerospace
Corporation concerning argon laser pumped CW lasing of Nd:YVO^ at
1 . 06 and 1. 34 [Jm.
2
Table 1. Comparison of Neodymium Laser Hosts
E/>
c:
o
cn
3
U
C
T)
£ .2
to 2
O ?
•—I '
c ffl
O'
f-H
1
-
-
CO
c
o
o
o
k
f-H
l-H
l-H
£
lO
lO
LO
— *
00
O'
X
X
X
a
•t
o
xO
r-
*-4->
V
xO
O
o
o
00
O'
1 u
U
o
•
•
•
•
•
•
l 03
o
•
l-H
f-H
l-H
in
O 00
CO
i a
u
o
r-
TJ
8
CO
00
xO
o
I
o
Ln sO
in
00
m
0)
<u
vh
I
0)
o
u
o
O'
O vO
• •
o o
a
O'
T>
6
oo
xO
O
X
O xO
00
IT)
xO
o
£
<D
u
o
u
o
00 00
O r-H
• •
o o
A
0 I
> <;
>■<
o
xO
o
oo r-
O'
X
o
m
ir» £
o
0)
0;
Jh
vm
<D
Jh
o
u
xO
o
o
m
o
5
s
n
3
Comments High gain, low Standard ^ matches glass. Low gain, C-axis
threshold optical quality. High gain, B-axis
size
II. MATERIALS GROWTH AND PREPARATION
A. Growth of YVO^ Crystals at Union Carbide Corp.
1. Background
The growth of YVO^ single crystals began at Union Carbide Crystals
Products in 1965 and small pure and doped crystals were grown by the
Czochralski process [4, 5]. The crystals were grown on a flame-heated
growth station from an iridium crucible. These crystals were supplied to
various laboratories including RCA and Bell Telephone Laboratory. More
recent work was funded by the U. S. Army Electronics Command at Ft.
Monmouth under Contract DA A B0772-C-0022. [3] This work included Bridgman
and Czochralski growth and was aimed at undoped YVO^ for polarizing prisms.
The technique settled on was RF induction heated Czochralski and boules of
1/2M dia. x 4" long were eventually produced. High optical quality polarizing
prisms were cut from the material. Following this work, some in-house
effort was put forth to grow some YVO^ crystals doped with Nd and with Er,
Tm, and H^. These crystals were, in general, small and the work was con-
cerned with growth habits, inclusions and melt composition.
The work of this report was begun under sub-contract to the University
of Southern California and was directed toward supplying 3 x 30 mm laser rods
and spectrographic samples of rare earth doped YVO^, ErVO^ and TmVO^.
2. Melt Effects
The phase diagram for Y O -V O (Figure 1) shows that YVO melts
2 3 2 5 4
congruently at 1810°C. It appears, however, that the melt is not stable under
normal furnace atmosphere conditions and that oxygen is lost from the melt.
At the growth temperature, the melt is very turbulent and complex dark con-
vection lines are observed. As the temperature is lowered, the melt forms
a crust of solid material on the surface. This crust can be removed by using
4
Temperature (°C)
1
Figure 1. Phase Diagram-System Y203_V2°s* After E. M. Levin,
J. Am. Cer. Soc. 50 [7] 381 (1967).
5
argon or nitrogen flow in the growth system. Liquid V O slightly above the
Z 5
melting point dissociates according to
v2°5
V2>5-x +
- o
2 2
Where more and more suboxides appear with increasing temperature. It is
believed that YVO, behaves in a similar manner. That is
4
YVO . -► YVO, + ~ O .
4 4-x 2 2
44
It has been shown that the concentration of V can be controlled for certain
practical oxygen partial pressures and that Czochralski growth can be effected
from the solution. While control of crystal diameter and atmosphere is criti-
cal and difficult, high quality boules have on occasion been produced.
Invariably, melts which have been solidified after growth are dark
indicating a reduced state and appear to have many phases present. Figure
2 shows remainders of melts from various growths.
3. Growth
During this research a total of 50 crystal-growth runs were made.
Growth was performed in a standard Czochralski growth station using a
2M x 2" iridium crucible. Pull rates were as high as 3 mm/hr with 1. 5mm/hr
settled on as it appears that inclusions are reduced. Various starting mate-
rials were used and availability is sometimes a problem. A list of available
materials is given in Table 2.
Initial growth runs on a new material were made with lower purity
materials while runs for yield were made with high purity ones. No effect on
crystal growth was seen as a result of purity. For production of samples and
laser rods the preferred material is General Electric's YVO^ powder of
99. 99% purity.
6
2. V„ Mi Y\/<J,|
3- Yi'-<I7
40C Yl fOfl XT t*n «i
B-Yi/-43
T
3% wJ:V^04»
On»uh 0«o#D *>\* LT Prta*'
Jtat» lr X.
Figure 2. Remnants of YVO^ Melts.
7
Table 2. Starting Materials Available for Crystal Growth
Material
Purity
Suppliers
Price
YVO .
4
99. 99%
General Electrics
$101. 60/lb
Tm2°3
99. 99%
Research Chemicals
4. 00/gm
Tm °
99.9 %
Research Chemicals
2. 75 /gm
99. 999%
Research Chemicals
200.00/lb
Er2°3
99.9 %
Research Chemicals
45. 00/lb
H°2°3
99.9 %
Research Chemicals
0. 35/gm
V2°3
99. 99 %
Research Chemicals
14. 50/lb
Growth was begun with stoichiometric melts at 60 standard cubic feet
per hour, scf/h, n^ flow with 1% atmosphere. A rotation rate of 28 rpm
and a pull rate of 1. 5 mm/hr were established. A seed of a-axis orientation
was used for most growths. Several growths were made off axis due to an
error in seed orientation. The Laue photographs used for orienting the seeds
for a-axis were actually (110). Laue patterns for each one are shown in
Figure 3.
The crystals exhibit a marked habit when grown along the a-axis. As
shown in Figure 4, the boule cross-section is flattened with faceting on (100)
and the long edge parallel to (001). The aspect ratio of the flattening can be
affected by pull rate. C-axis material, on the other hand, grows with a
nearly square cross-section.
Diameter control is very sensitive and difficult and frequent cross-
section steps were encountered. Figure 5 shows several of the shapes ob-
served. Some improvement in this condition was made by tuning control
equipment as closely as possible during the contact. As these changes occur,
the growth rate at the interface changes and inclusions of bubbles, second
phase material and occasionally iridium are observed to form on the inter-
face.
8
9
Figure 3. Laue patterns for (100) and (110) YVO
<0I0>
(b) a -ax is
Figure 4.
Cross-section Habits
of C-axis and
A-axis YVO
4*
10
1
11
Since the interface is nearly flat in YVO^, this means that inclu-
sions occur nearly along the a-plane. The a-plane is a strong cleavage
plane, but has sufficient strength to withstand normal fabrication stresses*
However, when inclusions are present on the interface the bonding is weak-
ened remarkably and cleavage frequently occurs in cool-down or in fabri-
cation. This is a major problem which must be solved.
Grown boules are in a reduced state and are dark in color. Sub-
sequent annealing in oxygen at 1400°C for 24 hours serves to return them
to transparency. No color changes were observed in the boules in normal
storage and handling.
Crystals grown were:
Material Growth Runs Yield of 4cm Boules
Nd: Y VO
4
' ’ABC": YVO
4
Er VO
4
Tm: Y VO
4
26
21
2
1
5
13
1
0
A list of actual compositions appears in Table 3.
4. Fabrication
YVO^ has a hardness of 480 KHN which is the range of many glasses
and as such polishes, saws, and grinds well. The major problem is that
of cleavage along the a-plane. It is felt that the problem may be due largely
to the presence of growth imperfections on the interface (a-plane) as already
discussed. Credence is given this hypothesis by the fact that laser rods do
not cleave into an infinite number of platelets. When a rod does cleave, it
usually withstands all subsequent grinding and processes without further
cleaving.
Because of high losses in grinding laser rod cylinders, core drilling
was investigated. Although core drilling provides a more gentle technique,
losses continued at the same rate. The problems in fabrication are
12
Table 3. List of Boules Grown
Table 3. List of Boules Grown (continued)
illustrated by the fact that crystals were submitted for thirteen 3x30 mm
laser rods and ten 3 x 20-30 mm while no rods were fabricated without break-
age and at least some loss of lengths.
5. Analysis
A cleaved sample of (Y -Er ^cTrnn )VO A was sent to
Seal Laboratories for analysis with particular attention to inclusions. Analy-
sis was done using scanning electron microprobe and electron microprobe.
Energy dispersive x-ray analysis (DXA) was used for the electron micro-
probe. This technique is useful from the surface to a depth of one micron.
A general area of the surface of the crystal. Figure 6, shows an
included particle (A), the general crystal matrix (B), and several loose con-
taminant particles. EDXA was performed on the particle at A, Figure 7 and
indicated the presence of major amounts of vanadium and erbium, a minor
amount of yttrium, and traces of thulium and silicon. The EDXA of the matrix
(B) indicated the presence of the same composition, except that no silicon
was present, Figure 8.
Included particles along a cleavage step is shown in Figure 9. EDXA
of the particle at "C" showed major amounts of erbium, vanadium and cal-
cium, minor amounts of silicon, yttrium and potassium and trace amounts of
sulfur, chlorine, and thulium (Figure 10), The EDXA of the adjacent matrix
(Figure 11) showed the same composition as the previous matrix.
A large inclusion. Figure 12, was also EDXA, the inclusion contained
major amounts of vanadium, erbium, silicon, chlorine, sulfur, potassium
and calcium, a minor amount of yttrium, and a trace of thulium (Figure 13).
Samples of the pure powders used to make the crystals were also
analyzed using EDXA (Figures 14-16). Using the peak heights and calculating
the number of X-ray peak counts per atom percent, the atom ratios of the
vanadate crystal were calculated: V:Er (plus Ho and Tm):Y were 2. 6:2. 7:0. 13.
Figure 6.
Included Particle (A) and General Crystal
Matrix in <YQ. ,?Er0_ 0?Ho0. 0] >V04.
16
17
X-ray Photon Energy (keV)
Figure 7. E.D. X. A. of area nA" in Figure 6.
>
<D
JX.
>N
CP
k_
Q)
C
LU
c
o
o
JZ
CL
>N
0
k_
1
X
18
Figure 8. E. D. X*A. of area "BH in Figure 6
Figure 9.
Included Particles Along a Cleavage
S'ep in (Y0_)7Er0_75Tm0_ 07H«0^,)VO4
19
20
Figure 10. E.D. X.A. of area "C" in Figure
(sjunoo 000*8 :9|dos nnj)
s^unoo uojoiy <dj-x |ojox
21
a;
Jh
3
w
E
-ray Photon Energy (keV)
SEAL
Figure 12.
A
(Y
Large Inclusion in
- , Er^ nrTm„
0. 17 0. 75 0. 07
HO0.01)VO4-
22
23
X-ray Photon Energy (keV)
Figure 13. E.D. X. A. of area "E" in Figure 1Z,
24
Figure 14. E. D. X. A. of pure VO powder.
-
. o
sjunoo uo|oiy ^dj-x |D40j_
25
X-ray Photon Energy (ke V)
Figure 15. E. D. X.A. of pure Er O powder.
(siunoo DO|/ooO'S9 :a|DDS||nj) o
s;unoQ uojoiy Adj-x |djoj_
26
-ray Photon Energy (keV)
Figure 16. E. D. X. A. of pure YVO powder.
B. Evaluation of Scattering Centers in Czochralski Grown YVO
4
Two single crystals of YVO^ grown by Union Carbide Corp. were
examined at Aerospace Corp. with a polarizing microscope to see if the cen-
ters responsible for laser scattiring could be identified. The crystals were
both Nd^O^ doped and were delibered from Union Carbide to USC at different
times.
One of the crystals was in the form of an MaM axis laser rod. Prelimi-
nary examination showed several zones of intense light scattering alont the
length of the crystal. Since this crystal was a reject due to the scattering, it
was decided to examine it more closely bo optical microscope techniques.
Two polished sections ere prepared, one parallel and the other perpendicular
to the rod axis. Each section was chosen such that it contained a volume of
the original rod that scattered light. Optical examination showed a high volume
of scattering centers such as those shown in Figure 17.
High magnification of these scattering centers showed them to have two
characteristics as shown in Figure 18a, b. The photomicrographs are taken
of the "a" axis slice of the rod. The difference of Figure 18a, b is that each
was taken with the polarizer parallel to each of the polarization directions of
the (100) plane. For one of the polarization directions the scattering centers
appear to be round black dots approximately 4 or 5 microns in diameter. For
the other polarization the scattering center appears to be more comples. The
rod shaped sharp index change appears around the dot and is aligned in ran-
dom directions in the crystal. Normally one would assign apparent index
changes to diffraction effects around the scattering center. In this case there
appears to be a sharp-irregular index change associated with most but not all
of the scattering centers. Due to the small size of the scattering centers, one
cannot clearly state whether the black dot represents a solid phase or a void
in the crystal. However, at this time a best guess would be that it is a void
and the endex change represents a rod shaped unknown phase associated with
the void. In-as-much as their distribution was not uniform along the length of
the laser rod, one can state that the crystal growth was not sufficiently controlled.
27
Spill
■■
%•
♦
*** 4
•«
Figure 17.
Scattering centers in Nd:YVO .
4
28
Figure 18. Scattering centers in Nd:YVO^.
The polarization of the illuminating light was
parallel to one of the polarization directions of
(100) plane in (a) and to the other in plane (b).
29
The other crystal was in the form of a quarter inch polished cube.
Microscopic examination revealed that the extinctions in the naM axis
direction were nearly complete and fairly sharp while that in the ncM axis
direction was incomplete and poor. The extinction in the McM direction
was light grey for all orientations indicating a McM axis wander or a constant
random strain that varied throughout the length of the crystal. Two types of
randomly distributed scattering centers were observed in this crystal. One
is of the type shown in Figure 19. The black triangular solid is very
thin, absorbs all light, and reflects metalic with incident light. Without
either chemical or crystallographic analysis one can only suggest its nature.
Generally, it fits the description of a metal such as Ir or Pt . This type of
included solid has been found in a number of high melting temperature oxides
grown from either Ir or Pt crucibles. Presumably the metal is soluable
in the melt, precipitates in the melt and is trapped by the growing crystal.
The size varies from 5 to ^15 microns. The other scattering center observed
was an extremely small black dot such as those shown in the photomicrograph
The size is 1 to 2 microns and as such is at the resolving limit of the micro-
scope. Both of these defects are randomly distributed over the crystal in
zones again indicating a lack of control during the crystal growth process.
30
Figure 19.
Triangular scattering center in
Nd:YVC>4. Possibly Ir or Pt.
31
C. Flux Growth of YVO^ at Aerospace Corporation
Single crystals of YVO^ and (Er,Y)VO^ were grown by standard flux
growth techniques in order to compare the defect properties of the crystals
grown at low temperatures with those grown by Union Carbide by melt
techniques. A small number of crystal growth runs were required in order
to produce the necessary crystals. The crystals were checked with a
polarizing microscope to determine the structure. They are optically uniaxial
with the same optical properties as tetragonal YVO^.
The crystals of YVO^ and Rare Earth doped YVO^ were grown from a
V^Oj- flux by standard flux growth techniques. They were grown from a
solution supersaturated by both slow cooling and flux evaporation. The
chemicals used for the growth of YVO^ were American Potash code 116
and J. T. Baker Co. Reagent Grade V^O^.
Standard form 50 mil platinum crucibles were filled with the appropriate
powder mixtures, their lids tightly crimped, and placed in a Super -Kanthal
heated muffler furnace. They were heated to 1250° C soaked 4-8 hours, pro-
gramed cooled at a rate of approximately 4°C/hr, then removed from the
furnace and allowed to cool to room temperature. The crystals were removed
from the crucibles by soaking in hot dilute HNO^ or HC1. The melt com-
positions used varied from 4 to 10 mole % with remainder being V^O^..
The crystals nucleated throughout the melt and varied in size from 0. 5 - 3 mm.
They were of two diverse habits (plate -like and equidimensional) with the
largest being of the plate type. The dimensions of the plate-like crystals were
approximately 3 by 4 mm with a thickness that varied from 0. 1 to 0. 5 mm.
The equidimensional crystals were generally =*0. 5 mm with a few as large
as 2 mm. No significant change in habit was detected with up to 10% of the
Y2O3 being replaced by rare earth ions.
The crystals varied from colorless to pale yellow from run to run with no
indication of any direct cause. It is felt that this was due to impurities in
the melt as little care was taken to insure high purity growth materials or to
eliminate introduction of impurities into the melt. Some of the larger plate -
like crystals had ordered trapped flux inclusions parallel to the major growth
32
planes. This type of defect can generally be minimized by judicious choice
of the growth parameters. Other than these defects, the crystals appear to
be high quality with no visible index changes or strain birefringence.
The basic habit of the crystals is that of {100} prism faces modified
by {lOl} pyramid faces (see Figures 20 and 21). The small equidimensional
crystals show a good tetragonal symmetry with all of the faces being approxi-
mately equal in size. The plate-like crystals, on the other hand, show a
decided lowering of symmetry to that of a orthorhombic crystal. The gross
variation of habit, i. e. , from equidimensional to plate-like, can be related
either to the growth temperature or to some significant change in the growth
mechanisms of the crystal. The small equidimensional crystals clearly grew
throughout the melt when the crucibles were removed from the furnaces and
as such grew rapidly. The plate-like crystals grew at higher temperatures
throughout the program cycle of the furnace and consequently at a much slower
growth rate. Such changes in habit are generally assigned to some major
changes in the crystal growth kinetics such as growth mechanisms, super-
saturation, etc. In addition, changes of this type normally involve the intro-
duction of new faces or a major reordering of growth rates the existing faces.
This particular case is somewhat unusual in that all the faces are of the same
type and the change of morphologic symmetry primarily related to a surface
energy change of the (100) and (010) planes. This is unexpected for tetragonal
crystals where one would anticipate variations only in the c/a lengths of
the crystals dependent on some kinetic change of the growth system. In the
absence of twinning or some recognizable growth feature as being responsible
for such a change in habit one must look elsewhere for its cause. It is un-
likely that impurities could affect the habit in this way for the (100) and (010)
planes in the tetragonal crystal system are equivalent planes and should not
be affected differently. About the only remaining cause is a structural one.
If so, and the full morphologic symmetry is applied then the crystal must
have existed at the growth temperature with a lower symmetry than tetragonal.
The full morphological description by faces type being (100) » (010) with the
( f 00) > (100) and the (011) = (011) with (101) » (101). One striking feature of
33
o
o
34
Figure 21. Flux Grown YVO^ Crystals
35
the crystals is that the (101) is consistently much smaller than the (101) or
is absent. This implies that the crystal had an orthorhombic distortion
with a^ ^2 anc^ a *oss symmetry along with a axis with a result that the
(100) and (100) as well as the (101) and (101) planes becoming unequivalent.
Optically, the crystal appears to be uniaxial as in the case for zircon.
36
III. SPECTROSCOPY OF RARE EARTH IONS IN YVO,
4
A. Introduction
The present interest in rare-earth-doped yttrium orthovanadate
(YVO^) stems from recent studies [6, 7] which showed that an experimental
laser rod of NdtYVO^ in a pulsed-lamp laser configuration had a threshold
at 1. 06 Mm about a factor of two lower than a standard quality Nd:YAG laser
rod in the same configuration. Furthermore, the performance of the vana-
date material in a CW laser configuration was comparable to the garnet even
though the vanadate sample had larger scattering losses [8], These observa-
tions arise out of the fact that thw 1,06 urn laser cross section in YVO, is
4
4*6 times greater than that of NdrYAG. In addition, the 1.34 (im operation
of the NdrYVO^ laser completely outperformed that of NdrYAG at 1. 32 Mm
because of an 18 times larger laser cross section of the vanadate relative
to the garnet [8] .
The first study cf a laser using NdrYVO^ was made in 1966 by
O’ Connor [ 1 ] who reported a pulsed-laser threshold at 1 . 06 [xm of ~ 1 J.
Difficulties in crystal growth prevented this work from progressing. How-
ever, recent advances in the growth of this crystal [3] have shown that
YVO^ is now a promising laser material. This is especially so because of
its very desirable mechanical, optical, and physical properties. For example,
it has excellent optical finishing properties because of its glass-like hard-
ness, it has only a slight tendency to cleave, and it has a high laser-induced
damage threshold. Since it is a strongly birefringent uniaxial crystal, the
dopant transitions are polarized; the lasing transitions are strongly polar-
ized parallel (n -polarized) to the optic axis (c-axis). The strong bire-
fringence makes possible electrooptic Q-switching without an intracavity
polarizer [6, 7].
Further studies of NdrYVO^ were described in 1968 by Bagdasarov
37
et al. [9] and in 1 969 by Kaminskii et al. [10]. These papers reported not
only a low threshold for lasing (~2 J) using this material, but they also gave
3 +
spectroscopic data for transitions of Nd at 4.2,77, and 300°K. The work
of DeShazer et al. [6] and of Bass et^ a_l . [7] presented energy levels, line-
4 4
widths, and peak cross-sections for the F , -* I . fluorescence transi-
4 4 ^ ^ 11/2
tions and energy levels of the F3/2^ !9/2 transitions all at room tem-
perature as well as laser performance data. In addition, unpublished spec-
troscopic studies by Pressley et ah [ 1 1 ] and by Karayianis et al, [l 2] have
been circulated wherein attempts were made to identify the energy level
scheme, particularly of the above mentioned manifolds. Among the studies
of these four groups, there is (1) only partial agreement as to the positions
4
of the Stark levels in the I ground level, (2) no agreement on the group
theoretical identity of the levels, and (3) no explanation of the absence of
4 4
certain lines from the ^9/2 transit^on grouP« For example,
Bagdasar ov et ah [9] and Kaminskii et al. [ 1 0] did not report on polarized
spectra or any level identities. On the other hand, Karayianis et al . [12]
and Bass et^ al . [7] both report polarized spectra, but their level identities
do not agree.
4 4 2
In the work described here, the Ig/2 ^3/2 anc^ /2 P°^arizec^
4 4 4
absorption spectra and the ^9/2 anc^ ^11/2 P°^ar^zec^ fluorescence
spectra, recorded at ~85°K and room temperature, are presented and
analyzed. We show that the spectra provides for unambiguous level assign-
ments. By examining the details of crystal field theory and the Judd-Ofelt
[13] theory of induced electric-dipole transitions as applied to Nd:YVO^,
we offer some possible explanations for the observed relative line strengths.
B. Experiment
The samples of Nd:YVO^ studied were obtained from two batches of
crystals grown by Crystal Products Department, Union Carbide Corporation,
38
San Diego, California* The first batch was produced by Dess and Bolin
[ 5 ] in 1 967-68, and the second batch by Rothrock and Wilder [14] in 1973-
74. Although the crystal from the first batch had a greenish tint added to
the usual sky blue color, the spectra were essentially the same. This
. 3 +
green tint was produced by color centers in the crystal. The Nd concen-
tration was nominally 1 atomic percent.
The absorption measurements were made usually by illuminating the
crystal transversely to the c-axis with parallel light from a qua rtz -iodine
tungsten lamp. This allowed the Gian- Thompson polarizer to be set either
parallel (n -polarization) or perpendicular (a -polarization) to the c-axis. The
experiment is shown schematically in Figure 22.
The spectrograph used was a 1 -meter Czerny- Turner mounting fitted
with an uncooled 7102 photomultiplier which was connected to a lock-in ampli-
fier and a log- converter (the latter being used only with absorption spectra).
The linear reciprocal dispersion of the spectrograph was about ]6A/mm in
first order. Corning glass filters were employed as needed to remove
unwanted orders and to avoid excitation of fluorescence transitions during
the absorption measurements.
The fluorescence experiments were performed by focusing the beam
from an argon- la s er -pumped Spectra-Phy sics dye laser along the direction
orthogonal to both the c-axis and the observation axis. The dye laser polari-
zation was parallel to the c-axis and the maximum fluorescence output was
obtained with the excitation set at about 583 nm. The dye laser power was
about 60 mW incident in the crystal.
The spectrograph was made insensitive to the incoming polarization
by inserting a calcite wedge in front of the slit with the optic axis set at 45°
to the slit. With the quartz-iodine lamp, we observed about a 2% change in
the recorded signal for a 180° rotation of the polarizer with respect to the
polarization requires that the illumination of the slit be reasonably uniform
along the height of the slit and that the image on the slit cover at least a
39
ABSORPTION EXPERIMENT
cr
Ld ^
to
CM
7 (T
GO M
cl cn
Z2<
<0-1
-1X0
CD H Q-
40
Figure 22. Experimental arrangements for absorption spectroscopy.
few millimeters of the slit height.
Wavelength calibration was accomplished by using Osram mercury
or neon gas discharge lamps and superimposing the spectra from these
lamps on the experimental spectra by a partially reflecting mirror on the
optical axis of the spectrograph entrance slit.
C. Spectra
Since out primary objective was to clarify the positions and identities
4 4 4
of the levels of the *9/2 ground state and of the ^j/2 anc* F3/2 states of
interest in laser applications, we emphasize the better resolved spectra
taken at ~85°K; however, for completeness we also present data taken at
room temperature.
The absorption and fluorescence spectra used in the analysis is shown
in Figures 23 through 26. All these spectra shown were taken using a sam-
ple 7. 8 mm thick taken from Boule 3L of Rothrock and Wilder. Absorption
spectra taken at both temperatures with boule 3L samples and the Bolin-
Dess samples in longitudinal illumination (i. e. , along the c-axis and with
no polarizer) showed excellent agreement with the corresponding a spectra
shown in the figures, both in line positions and relative strengths. The
strengths of these longitudinal a-spectra were a factor of two larger than
their corresponding transverse a-spectra as is expected from the simple
theory of the radiation pattern of a a transition. On the other hand, as can
be seen in the figures, the tt -to-a strength ratios do not follow any simple
relationship. These ratios can be understood only after a careful examina-
tion of the J composition of the crystal field levels as it influences the transi-
tion matrix elements obtained from the Judd-Ofelt theory [13].
41
ABSORBANCE t''-MI0/I ) (cm'1
4 2
Figure 23. Polarized absorption spectra of the 19/2"* ^1/2
tions of Nd:YV04 at ~300°K. The letters a, b, c,
4
denote the crystal field levels of the I9/2 state,
c-axis was transverse to the observation a&ifc.
transi-
and e
The
42
ABSORBANCE >%(!„/!) (cm")
4 4
Figure 24. Polarized absorption spectra of the I9 / 2 F3/2 transitions
of Nd:YVC>4 at two temperatures. The peak of the flat-topped
rr - line at ~ 300°K was determined to be at 18. 5 cm"',
Crystal oriented as in Figure 23.
43
300° K P ~85°K
(SlINfl AdVdliadV) A1ISN31NI 3AllV33d
<D
ro w
o>
^ v v o
bJD
U o ® T?
X > JC ®
fD >
H £
03
03
o
Ih
D
3
l
>
4->
<D
<D
>>
>*
03
Ih
JC
4->
XI
(U
Cl,
d>
d>
Z
G
JC
L_i
»*-i
c
d>
JC
r>
LT)
ro
<L>
in
D
W)
44
(SlINn AdVHliadV) A1ISN31NI 3AllV13a
c S
o .>
4J M U
« 0) (TJ *S
c ^ ^ t:
2 ^ ^
id cm ^
=*•* .•
t O ~ 01
fsj^ o
CM
ro Cu
CO
Ui ^ *0) i-T
<1)
<D
_C • r!
i U J 5
't* 0) 4) O
2 -
u *
& 2
w O
b£)
<D
-C
H
i be
* .s
+■> bf)
« C
£r °
u a;
o
_C co
4-> ~-<
0)
« >
4_i ^
o «
C ■"*
<D <D
TD -C
tJ •
(D CO
T>
n a; 5
^ *
d -
« *
Jh CO
a; >h
a a;
£ 5
Jh
Ctf
o
a
"U
<d
<D
o a>
* ^
u
X
<D
I
U
<D
CO
aJ
1 >
« Sh
>> V.
*0 r0
0) £
H o
c
0)
T3
CO
u •
a; m
(M
4-»
a; q;
TJ 3
a; bo
0) ju
H Eh
^ *H
^ d
a .5
§ CO
<tj
« _
-c ^
+j <D
4-»
<d a
^ <D
-g t
£ O
VO
CM
<D
U
d
bC
45
transitions
1 4i
U 9/2 *1/2
4 2
The spectra (taken in third order) for the 1^^ "* ^1/2
in Figure 23 are given only for room temperature since they show the transi-
4
tion originating from the highest level of the 1^^ man^°^. This transi-
tion is frozen out (i.e., the initial level is thermally depopulated) at ~85°K.
The spectral region of these transitions was carefully searched at both
temperatures and polarizations for additional lines corresponding to transi-
tions to the ground state levels other than the five shown in Figure 23. None
were found.
We do see in the n-polarization spectrum in Figure 23 a weak and
very broad band under the strong line. Similar bands are found in nearly all
spectra; yet, they were too weak, too broad, insufficient in number, and
not consistent in their observed pattern to be considered part of the electronic
3 +
transitions associated with the primary site population of the Nd ions.
2. 4I
9/2
3/2
Figures 24 and 25 show the absorption and fluorescence spectra for
4 4
the ^3/2 *rans lt'lons at the two temperatures. Because of the large
thickness of the sample used to obtain these spectra, the high- cross -section
transitions completely absorb the incident light thereby causing these lines
to become flat-topped. The large thickness was used to insure that all perti-
nent electronic transitions were found.
In fluorescence, the strong n transition was observed to be quickly
self-absorbed as the excitation beam was moved back from the sample sur-
face. Self-absorption means that, since these transitions terminate on the
ground manifold, the fluorescence is subject to absorption during passage
out of the sample. As seen in Figure 25, some residual self-absorption
remains at ~85°K even though some effort was made to place the beam to
within a tenth of a millimeter of the sample surface. An inset in the figure
46
shows how strong self-absorption occurs when the beam is about 1-2 mm
back from the surface.
In absorption, we note that the lowest wavelength line which appears
in a spectrum at ~85°K also appears in the corresponding tt spectrum as
well as in the ~85°K tt -fluorescence spectrum. This behavior is also
4 4
observed in the ^3/2 Ij ^ ^-fluorescence spectra shown in Figure 26
wherein two o lines appear on either side of the strong tt line in the low
temperature spectrum. These lines are not apparent in the corresponding
room temperature spectra because of interference from the neighboring
strong tt line. These observations may have been caused by polarizer or
sample misalignment. However, the lines appeared in independent experi-
ments on different samples. Thus, we are led to the possibility that the
3 +
selection rules are broken which could mean that the Nd site symmetry
is lower (i. e. , ) than the normal symmetry at that site.
The situation is complicated by the fact that all of our samples
showed strong and complicated strain patterns when viewed along the c-axis
between cross polarizers. These patterns would completely change with
only slight changes in the sample orientation. When viewed perpendicular
to the c-axis, comparatively few patterns were observed with good extinc-
tion being achievable at the appropriate orientations. Thus, it is possible
that the anomoly described above is produced by crystal imperfections such
as would produce wandering of the c-axis through the crystal.
Bagdasarov [9] reported two sites which they designed I and II.
Our results are in good agreement with their Stark level positions for
type I sites. We do not observe any lines which we can identify with type II
sites. In addition, they reported weak lines on the long wavelength wings
4 4
of the strong 0 lines in the ^3/2"* ^11/2 ^uorescence spectrum at 77°K
which they called "companions. M We see similar lines on the strong 0
lines in the absorption spectrum for ~85°K shown in Figure 24. These
lines are spaced about 1 cm * less than their strong neighbors. The
47
sharpness of these weak lines suggests that they are from a well defined
3 +
Nd site, and their smaller spacing indicates a slightly weaker crystal
field. Examining the crystallographic data given by Wyckoff [1 5J for crystals
having the zircon structure illustrated in Figure 27, we calculate a yttrium-
o
oxygen nearest-neighbor spacing in YVO to be 2. 296 A . Using the
o I o | ^ ^
following radii, Nd -1.08 A, Y - 0.93 A, and O -1.40 A, we find the Y-O
3-I-
near est- neighbor spacing of the Y site to be 0. 184A smaller than the
34 o 34
radius of the Nd ion as compared to a 0.034 A mismatch for the Y ion.
34
The mismatch of Nd is certainly sufficient to introduce the possibility of
a site arising from a slight well-defined distortion of the basic symmetry
3+ 2d
normally present at the Y site
3‘ 4F3/2Al/2
The fluorescence spectra at room temperature and ~85°K for the
4 4
^2/Z^ *11/2 ^rans^*lons are g’lven in Figure 26. After group theory there
should be 1 2 a and 6 tt lines. Only the low temperature a spectrum shows
the required 12 lines. In the tt spectrum, neglecting the 2 anomolous G
lines discussed above, we have only 4 lines. We propose that these two
anomalous G lines are not the two missing tt lines. In support of this claim
is the fact that in all the other spectra, a TT line is never significantly in-
creased in strength by changing to G polarization contrary to the behavior of
the anomolous G lines. The strong TT line appears in a polarization as the
second of the four strong lines, counting from the short wavelength end. We
shall identify the third line of this group as being one of the missing TT lines.
This line can be seen on the long-wavelength side of the strong tt line only
as an asymmetry in the lower half of the strong TT line.
The spectra given in Figure 26 for ~85°K was not taken with sufficient
gain to reveal the "companion” lines discussed in the previous subsection.
With higher sensitivity, we observe the companion lines reported by
4 4
Bagdasarov L 9 ] for the F^^"4 Ij ^ transitions.
48
D. Analysis
1. Site Symmetry
i structure of YVO
to the space group D„u [15]. The Y"51^ site symmetry in which the Nd
The structure of YVO^ is that of zircon which is tetragonal and belongs
4h
,3+ .
ions are found is D_,« Referring to Figure 27, the Y ion is at the cen-
2d 2-
ter of a tetrahedron formed by four nn O ions. The reduction to
symmetry arises from the fact that this tetrahedron is shorter along the
c-axis than the corresponding dimensions perpendicular to the c-axis.
3 +
In addition, this irregular tetrahedron is surrounded by four nn Y ions
2-
and their O tetrahedra which are located below, left and right, and
above, front and back, of the site in question. These four Y-O tetrahedra
form a larger irregular tetrahedron at the center of which is the site in
question. This large tetrahedron is oriented 90° from the orientation of
5+ 3 +
the Y-O tetrahetron at its center. The V ions are not seen by the Y
5 +
ions in the first approximation since each V ion, due to its small size
t 2-
(0. 59A ) is completely shielded by a tetrahedron of O ions.
This intermingling of tetrahedra produces two kinds of equivalent
Y (hereafter the superscript for charge is suppressed) sites related to
each other by a 90° rotation about the c-axis. It is possible that the mis-
match between Nd and the Y site described in the previous section could
remove the equivalence of a fraction of these two sets of sites thereby
giving rise to another set of sharp lines.
We also note that neodymium ions located in nn Y sites would be
expected to see a significantly lower symmetry and larger crystal fields.
It is possible that the lines identified as site II by Bagdasarov [9 ]
are from Nd ions in double clusters since their spectra were taken with
samples having about twice the Nd concentrations as our samples which
makes the concentration of double clusters four times larger in their
samples compared to ours.
49
o<
v a>
O'
tL *>
>* . u
w rt
0 V
rj rd
o ^
1 r- V
-£—* nj
Q £ *8
"— a;
— t xju:
<D L
U r> -4->
+> ^ £
c
a
rt
C
o
W>
co JS
s s
O «J
w X
c o
I
£
a
a>
J3
H
c
8 xi
, rt
zd C
a> crj
a >
6
rt
c
rt
>
b-
(NJ
a>
W)
£
50
have been omitted from the cell to enhance clarity.
2. Electric Dipole Transitions Between Crystal-Field Levels
The importance of Nd:YVO^ is in the larger cross section and the
corresponding higher gain of the laser transition compared to the laser
transition in Nd:YAG [6,7], A n understanding of this difference can be
obtained by analyzing how the crystal fields of the two hosts influence the
2S + 1
dipole oscillator strengths of transitions between the L manifolds of
j
rare earth ions in crystals can be expressed as
fT.Tz = X VT (2J + if1 |< [tSL]j|| U(t)|| [t'SL']J'>|2 (1)
J J t=2, 4, 6
where V is the center-of-gravity frequency of the transition group, the parame-
ters includes quantities which characterize the opposite-parity configu-
rations and the strength and symmetry of the crystal field, the reduced
matrix elements are of the unit tensor calculated in the intermediate
coupling scheme, and the remaining symbols have the usual meanings.
The usual procedure is to fit the parameters using experimental oscil-
lator strengths. Unfortunately, this approach does not predict
the oscillator strengths of transitions between specific crystal field levels.
In fact, the parameters cannot be used to determine these strengths
since the scheme by which the available oscillator strength is portioned
out to the symmetry-allowed transitions is directly determined by the
specific distribution and the type of ions surrounding the rare-earth ion
and the J content of the crystal-field levels (which is also directly
z
related to the surrounding environment of ions), while both of these details
are removed from T parameters by summations over all possible values.
As it will be shown, it is these details of the crystal field interaction with
Nd that make the YVO, host of special interest. To make this evident, we
4
shall draw on comparisons with YAG and CaWO^.
51
Equation (1) is obtained from a more general expression of the Judd-
Ofelt [l 3] theory which includes an accounting of the details of the crystal
field and the J mixing. A useful form of this expression for the oscillator
z
strength of a transition between two crystal field levels a and 0 for polari-
zation orientation ? is given by
f = V
a-»pe a0
+ !>«;<*'>£ s'(k,t)£ (-d’ +Pe (,)V
q,P
.(t)
q+P\p -(q+p) q
t= 2,4,6, k = 1 , 3, 5, 7, |q|<k, p= 0, +1,
(2)
where V is the transition frequency, the ( U^) are the matrix elements
a8
which are the same as in Eq. (1), the H (k, t) are parameters which contain
quantities that characterize the opposite-parity configurations which are
responsible for inducing the electric dipole transitions within the 4f con-
(])*
figuration and other constants, and the e are the complex conjugates
P
of the polarizer components in tensor form where p = 0 corresponds
to rr transitions and p= 1 1 are for a transitions. The A are the parame-
q
ters of the odd-parity terms (i. e. , k odd) in the usual multipole expansion
of the crystal-field potential energy Vc. They are functionally dependent on
the coordinates of the surrounding charges. We note that these odd-
parity terms can only be non-zero in sites lacking inversion symmetry.
The quantities X p are g^ven
r(t)L « = E C* (J ) Cft (J ') (-1)
q+p |-£/ a Z p z
J-Jz / J t J'
\-Jz q + P K
(3)
where the Cs are the J -mixing coefficients and the %>' s are integers that
specify which J values are present in the corresponding states, and
52
therefore, which are to be considered in the summations of Eq. (3). The
values of the J^-mixing coefficients are obtained from the crystal field
calculation which requires detailed knowledge about the spectra. This cal-
culation involves products in Vc of the form
Vc *
(4)
where k = 2,4,6 only and the 3-j symbol is identically zero when the num-
bers in the top row cannot form a triangle or when the sum of numbers in
the bottom row is not zero. In the absence of any external fields, the states
of the ion are specified by a single value of J. However, the 3-j symbol
in Eq. (4) connects states having different J and J values. Thus J and
z
J are no longer good quantum numbers in the presence of a crystal field,
z
In the description of crystal-field splitting, Hellwege [16] showed that
for sites having a well-defined p-fold rotational symmetry axis, the J
z
composition of a crystal-field level is determined by a crystal-field quan-
tum number as given by
Jz=H + § p s p (mod p) (5)
where p is the fold symmetry of the axis of the crystal field (assuming only
non-cubic symmetries), and ? = 0, t 1 , t 2, • . • In NdrYAG, the Nd site
symmetry is with p = 2. However, in Nd:YVO^, the site symmetry is
D. which includes, in addition to three 2-fold rotational axes, a S sym-
2d 4
metry operation which is a pure rotation of 90° followed by a reflection
in a plane perpendicular to the axis of rotation about one of these axes
(i. e. , the z axis). As the result of this operation, Eq. (5) is no longer
sufficient to describe the decomposition of an arbitrary J state in a D
crystal field. This can be understood by determining the requirements
that the operation places on the crystal-field wave functions.
In general, the states associated with a crystal-field level a of a
53
given J manifold can be written as
,aM) =Zca(MJz) |JZ> ’
(6)
where the (|iJ ) are the same as in Eq. (3) and the summation is over
values of ? to be determined. The and operations are equivalent
to a 4-fold rotation followed by inversion as specified by IC and IC ,
f r
respectively. A p-fold rotation operation^ (2rr/p) followed by the inver-
sion operation J can be expressed as [15]
J lap) = S I a.U> » (7)
where^/ = £) ( 2 tt / p ) and the operations are defined as follows:
JMli
EAj
(x, y, z) = ill (-x, -y, -z) = (-1)1 ill (x,y, z),
and 2ttJ
, z
JD ( 2 tt / p) 'I'j (r, P, Cp) = ijij (r, 0,<P + 2n/p) = el P ijr j (r,0,cp),
z z z
th
where £. is the orbital angular momentum quantum number of the i electron
in the configuration of interest. Substituting Eq. ( 6 ) into Eq. (7 ) allows
the states to be separated into classes of states having a common phase
factor specified by which depends on whether the configuration has even
parity (i. e. , even) or odd parity (i. e. , ££. odd). If the configuration
i 1 i 1
is even, Eq. (5) is obtained. If the configuration is odd, as it is for Nd,
then we find for the Sp operation,
M= V 2 +?P-
Since £ LS a running index which can have either sign and has no physical
significance of its own, the two sign choices in the above expression gives
the same result for a given J 1 = J value. Therefore, we chose to use
z'max
54
J = (m + +“) (mod p) . (8)
z Z
The tables of Koster et al» [lTj show that in D symmetry, crystal-
3 2
field levels arising from a J state of the 4f configuration of Nd are of
only one irreducible representation (rep) namely T- with U - (i. e. ,
doubly degenerate). Whereas, in D symmetry, two doubly degenerate
Zd
reps are pos s ible, namely I\ with [x = +1/2 and T_ with (a = f 3 / 2. The J
o “ 7 z
values permitted in the crystal-field rep of the symmetry using
Eq. (5) with p = 2 are as follows:
For|j = +1/2, §= 0, 1, -1, 2, -2, 3, -3, ....
J = 1/2, 5/2, -3/2, 9/2, -7/2, 13/2, -11/2,
z
and for |J = -1/2, ? = 0, 1, -1, 2, -2, 3, -3, ...,
J = -1/2, +3/2, -5/2, +7/2, -9/2, 11/2, -13/2,...
z
Likewise for the Ty and reps of symmetry, we obtain the following
values for Nd from Eq. (8) with p = 4.
For|j=-3/2, ?=0, 1, -1, 2, -2, ...,
U= +3/2,
M= -1/2,
J = 1/2, 9/2, -7/2, 17/2, -15/2, ...,
z
? = 0, 1, -1, -2, -3, ....
J =7/2, 15/2, -1/2, -9/2, -17/2, ....
z
5=0, 1, -1, 2, -2,
J =3/2, 11/2, -5/2, 19/2, -13/2, ...,
z
and for |J= +1 /2, 5=0, 1, -1, -2, -3, ....
J = 5/2, 13/2, -3/2, -11/2, -19/2,
In symmetry, the free-ion states split as follows [17, 18]:
2
P , -» r
1 / 2 5
-+ 2T
3/2 5’
W5V
w-s-
55
In D2d symmetry, for 4f , an odd-parity configuration, we find
•p -» r
1/2 A7*
f -*r + r
3/2 67’
W*VSr7
I -4 3*p -i- 3p
11/2 6 x7 •
We note that in even-parity configurations, F^ and T exchange places in
the above list. Thus, in YAG, the two crystal-field levels belonging
4
to the state each can have a J composition consisting of a mixture of
3/2 z r °
J = t 1/2 and t 3/2. On the other hand, in
ind
I\(|i = + 1/2), pure J = 4 3/2,
o z
r_( n = + 3/2), pure J =4. 1/2.
Similarly, in Nd:YAG
for 4I r (M= ± 1/2), mixed J = ±1/2, 43/2, ±5/2, 47/2, +9/2,
V / Z 5 z
and for 4I . , rc( \l = ±1 /2), mixed J =±1/2, 43/2, ±5/2, 47/2, ±9/2, +11/2.
11/25 z
Whereas, in Nd:YVO„
for 4I
9/2’
and for I
r,(U= ±1/2), mixed J =+3/2, ±5/2,
o z
r7(U= ±3/2), mixed J = 4I/2, +7/2, 49/2,
f z
I\(|i = ±1/2), mixed J = +3/2, ±5/2, +11/2,
o z
r (H= ±3/2), mixed J =+1/2, +7/2, +9/2.
f z
It should be noted that from Eq. (4), the values permitted in a given
state cannot be changed by J mixing unless the symmetry is changed.
11/2’
56
The differences between these two symmetries are significant when
k
considering Eq. (2). In symmetry, the non-zero A parameters
occur for all possible odd values of k with |ql = 2,4,6 and *q < k. However,
3 5 7 7
in D symmetry, onlyA^, A A and A ^ are non-zero, in general.
The two 3 - j symbols that appear in Eqs. (2) and (3) govern which parame-
k ... 44
ters A are active in a given transition. Thus, for the *9/2^ ^ ^3/ Z
transitions, t = 4, 6 and k = 3, 5, 7. Since all permitted values of q can
occur in symmetry, then from the rule on the bottom row of the 3-j
symbol in Eq. (3), namely
J = J ' + 0 + q,
•7 7
(9)
it is clear that all transitions are permitted in both TT and a polarizations.
This is really trivial because with only one crystal field state, viz. IT, this
5
must follow logically.
In D symmetry, since q = + 2, +6 then we have from Eq. (9) that
Zd “
J = J / + 0 i 2, and J - J / f p + 6,
4 4
Using this relation, Table 4 is obtained for the *9/2 ** ^3/2 transltlons*
The significant fact obtained from Table 4 is that the transitions which
4
couple to the T crystal levels of the I . state depend on the contributions
3 5 7 7
to Eq. (2) of the ^ 42 * ^ +2 ^ +2 *errns relative to that of A ^ • Thus .
if one of the IT levels in that state had pure or nearly pure |J | = 9/2 compo-
7 z
sition and the A term was dominant, then the TT transition with the state
and the a transition with the T, j state would be expected to be stronger
than the a transition with the state. Also, a transition with a given level
having |j | = 3/2, 5/2 composition (i. e. , T^) would be comparatively weak
if not completely missing. On the other hand, if q was restricted to +2 due
7
to the smallness A , then
57
Table 4.
Transitions permitted by the 3-j symbol in Eq. (3) for the
4 4 3-1- -a
I q i 2^* ^3/2 ^ransitions of Nd in D2J symmetry.
J
0
j'
z
q - +2
z
q = ±6
0
+1 /2
+5/2, -3/2
- - 1
(r6)
TT '
-1/2
(r7)
0
-5/2, +3/2
“ “
r+i
+ 1/2
+7/2, -1/2
-9/2
cry
+i
-1/2
+ 5/2, -3/2
- -
<r6>
a 1
-1
+ 1/2
+3/2, -5/2
- -
<r6>
.-1
-1/2
+1/2, -7/2
+9/2
(r7)
0
+3/2
+7/2, -1/2
-9/2
TT '
(TV)
(T7)
0
-3/2
6 +1/2, -7/2
+9/2
'+1
+3/2
+9/2, +1/2
-7/2
<r7)
+1
-3/2
+3/2, -5/2
_ _
<r6)
a '
-l
+3/2
+ 5/2, -3/2
- -
<r6>
-3/2
-1/2, -9/2
+7/2
<r7»
a. See text.
58
r? (|jj = 9/2) n-> r6 and a-»r?
would not be observed or would be very weak.
4 4
Applyingthe above analys is to the F^^*— > *9/2 trans^lons» we would
expect to find 10 lines in the Nd:YAG spectrum which is the case f 1 7].
However, from Table 4, we have sufficient reason to expect lines to be
missing from the spectrum of Nd:YVO^. The symmetry allowed transitions
in the host are found from Eq. (8), where [i - - A|i = AJ (mod 4), which
z
becomes, from Eq. (9),
A(i = ( p + q) mod 4, q = f 2, f 6 . (10)
The selection rules which result from Eq. (10) are Ajj = +2 for tt polari-
zation and A)i ( P = +1 ) = 1 1 , +3 for a polarization and are summarized in
Table 5. Applying these rules to the transitions in question indicates we
should observe 10 lines in a polarization and 5 lines in TT polarization. The
low temperature fluorescence spectrum given in Figure 25 shows only 8a
lines and 3 tt lines with one tt line very strong supporting the implications
derived from Table 4,
3. The P t State in D Symmetry
1/2 2d
2
The 2-fold degeneracy of the Pj/2 s*'a*e' being of magnetic origin
(i. e. , spin) cannot be split by a crystal field. Furthermore, the spin
4
selection rule AS = 0 forbids transitions between the I ^ ground state and
a doublet. However, Rajnak [19] has shown that the "free-ion" eigen-
2
vector for the energy level associated with the P . state contains a signi-
4
ficant admixture of the ^\/2 s*a*e which will accept transitions from the
ground manifold. Thus, by observing these transitions, it should be possible
to unambiguously locate and identify the levels in the ground manifold. From
Table 5 , there can be 5 a lines and only 2 tt lines between 2l\ + 3T levels
o 7
59
Table 5.
Electric dipole selection rules
in D , Symmetry a
Zd
r6(i/2)
r?(3/2)
r6o/2)
a
rr a
r7(3/2)
tt a
a
a. The quantities in parentheses are the
crystal field quantum numbers |p|.
60
and one level. Figure 23 shows 2 rr lines but only 3 a lines. This clearly
fixes the two I\ levels in the ground manifold and implies that the ground
6
state is I* . The missing a lines can be assumed to be due, at least in part,
to the details of the J compositions of the levels associated with the missing
z
lines.
E. Energy Level Scheme
4 it 2
The spectra in Figures 23-25 show that one of the ^9/3 i/2
transitions originates from a level that also couples to one of the
4
F / levels in n polar ization. This means that choosing the label T_ for
J f L* ■
the 2p level must, if we follow Table 5, fix the order of levels in the
4 * ^
F3/2 state to be Fy ( = +3/2) high and T^([l = +1/2) low in energy. Since
the ground state couples to the lower level in tt polar ization, then the ground
state must be T ^ as found in Section D. 3. above. From EPR studies of
Nd:YVO^, Ranon [20] reported that the ground state could only be fitted to
a linear combination of the |j I values 1/2, 7/2, and 9/2. From Table 4,
1 z 1
this corresponds to a r level consistent with our findings.
As previously suggested [12], because of the similarity between the
4
crystal-field expansions for CaWO^ and YVO^, the direction of the FQ ^
splitting in the two hosts should be consistent with the sign A^. This comes
from Eq. (4) which, for J' = J = 3/2, only k = 2 gives a non-zero value.
2
The sign of the A parameter can easily be found from
E_ 2 _ , 3-, 2 2. . ;
r e Q./r. ] (3z. - r. ) / r.
J J J J J
J
.th .
where e is the electronic charge, -eQ. is the charge of the j ion, r. is the
J
.th .
J
radial separation between the site center and the j ion where z. is the
corresponding z component, and the summation is taken over all the ions
surrounding the site. If all the ions or groups of ions of the same kind are
located at the corners of regular tetrahedra having a common center at
61
£
which the Nd ion is located, then z. = +r. If 3 and A ^ = 0. In D _ symmetry,
j j 0 2d 7
as described in Section D. 1 • , the tetrahedra are shorter in the z direction
2 2
than in the x and y directions. Therefore, z. < rj /3 and since the net
charges of the surrounding ions are negative, A ^<0 as was reported for
Eu3+ in YV04 [21] and for Nd:YVC>4 [10, 12]. The sign of in CaWC>4 has
been reported to be positive [22] with the (j = ±3/2 level split down consistent
with our results for YVO,.
4
4
Having established the ordering of the states in the manifold
as described above, the identity of the four observed transitions in the
4 n 4
F ^ ±11/2 sPectrurn can be established. In the discussion concern-
ing the anomolous appearance of two of the strong a lines in the rr spectrum,
we concluded that there is a fifth TT line in the long wavelength wing of the
strong tt line. This is sufficient to complete the specification of the 6 levels
4
to be alternating H and II with the lowest level of the I . manifold being
67 11/2
rr
The energy level scheme based on the above deductions is given in
Figure 28 for room temperature. Also shown are the observed transitions.
The experimental frequencies of these transitions taken at room tempera-
ture and ~85°K are presented in Table 6. The line widths and peak cross
sections of the ^F^^ ^4^^i]/2 trans^lons a^ roorn temperature were also
measured. These results are given in Table 7. The strengths were deter-
mined by comparing the fluorescence intensities to the intensity of the
4 4 4 4
*9/2^ F3/2^ resonance line. The laser transition F^2(a) *11/2^
is at 10,640. 9 + 0. 2 A and is predom inantly tt polarized. It was found to have
a peak cross section of 4. 6 times greater than that of the 1. 064 jam laser
- 1 9 2
transition in Nd: YAG which has been found to be 6.5x10 cm f 9].
62
O- POLARIZATION
r7-
7 r POLARIZATION
23036 cm'1
r7-
r7-
r7-
r7-
437
201
174
112
11380
1 1 365
2180
2151
(2062.3)
■ 2046
. 1988
1968
Figure 28. Energy levels (+ 1 cm ) of selected states of Nd:YVO^
at room temperature showing transitions resolved at
~85°K. The approximate positions of the levels at~85°K
are shown as short dashed lines, (a) This is the ~85°K
value. See Table 6.
63
3-1-
Table 6. Observed Transitions of -1 Atomic % Nd in YVO,.
4
Transition
_ , . . b
Polarization
v .
air
~85°K
(cm- * )
~300°K
4T 2_
*9/2"* pi/2
a - a
a
23 040.4
23 036. 3
b - a
a
22 928. 9
22 922
c - a
a , tt
22 866. 1
22 863
d - a
-
n/o
n/ o
e - a
TT
n/o
22 600
4 4
F3/2 44 T9/2
a - a
a , tt
11 368. 1
1 1 365.4
a - b
a , tt
1 1 258. 6
11 253. 2
a - c
-
n/o
n/o
a - d
a
11 1 58. 2
~11 164
a - e
a
10 931
10 929C
b - a
a, (tt)
11 386. 1
11 379. 6
b - b
a
11 276.4
11 26 7C
b - c
a
11 211. 2
11 20 5. 5
b - d
_
n/o
n/o
b - e
a, tt
10 948
10 943
4 4
F3/2"* Ill/2
a - a
a , tt
9 400. 9
9 397. 7
a - b
a, (tt)
9 379. 9
9 377. 5
a - c
a , tt
9 320. 8
9 319. 1
a - d
a
9 306. 2
n/o
a - e
a , tt
9 213. 5
9 214. 8
a - f
a
9 186. 7
9 18 5. 9C
b - a
a, (tt)
9 418. 6
9 411. 9
b - b
a , nd
9 397. 7
9 391. 6C
b - c
a
9 338. 2
9 333. 7
b - d
a, nd
9 323. 5
n/o
b - e
a
9 230.6
9 228. 8
b - f
a, tt
9 203.8
9 200. 0
Values given to 0. 1 cm'* are accurate to + 0. 5 cm“^ or better.
All other values are to + 1 cm“*.
a. See Figures 1-4 for traces of spectra, n/o means ’’not observed. ”
b. The a transitions which appear weakly in tt polarization are denoted
by (TT).
c. These values had to be deduced from other values because of inter-
ference from neighboring lines.
d. This polarization was not resolved. See text.
64
Table 7. Parameters for F , -* I . Transitions of Nd: YVO at Room Temperature
o
o
o
(\T
£
O
O'
X
o3
i
o3
bC
C
O
C
O
*4->
0
<d
w
1
CD
CO
o
Sh
U
<d
ft
X
rt
i
o
bo
d
o
u
■S _p
* 'c
<D c
d
3
o
§ 7
cr £
<d o
a
o
to
d
rf
Jh
H
o
CO
sO
2
o
a
o
co
o
a
o
a
o
c
o
CO
r*
o3
<d
£
o
a
O'
O'
vO
oJ
O
o
CM
r*
CtJ
O
rX
Cti
•
1—1
CD
*
d
d
•
CD
*
•
CM
CD
*
rX
CM
X
vO
a;
•
4->
•
aJ
0
o
•
03
•
03
T3
X
CO
cm o r- cm
CD
-s*.
r-
CD
vO
CD
CD
CD
£
d
d
>
4->
00
in
r-
00
t-
•
r-
•
vO
•
vO
•
r-
•
CM
•
vO >—»
• •
-H O
CM
m
vO
in
f-H
in
CM
00
00
O
CT-
H
r— (
r-’
CO
Ch
co
vO
00
O
m
r—<
O
r-
CO
r— l
CM
O
CM
rH
o
00
CO
CO
CO
CO
co
co
co
CM
CM
CM
a-
a-
O'
o
O'
O
o
O
O
O
0s
03
03
rO
U
a
TJ
TJ
CD
CD
t
t
T
t
t
T
t
t
T
T
t
t
rO
03
03
r£>
03
rt
03
vO
CD
£
Ol
H
<D
CD
cn
Jh
<d
co
O
4->
O
c
to
d
03
<d
£
ctS
u
4->
o
CD
a,
to
to
CD
O
03
-M
O
u-i
vO
(M
<d
d
bo
<u
<d
co
u
a
<d
d
cr
<d
Jh
-C
bO
CD
*
T3
£
be
c
’>
-C
CD
«
TJ
C
OJ
CO
a)
d
rt
>
«
o
in
oo
*
<D
U
03
to
a)
CO
CD
-C
4->
d
<D
U
a
CL
<D
CO
a
£
03
co
0
m
1
CO
CD
<d
Q
£
O
U
Vm
T3
CD
c
rQ
O
<D
Jh
<d
*
Ti
03
T3
CD
co
CD
x:
H
o
(M
CO
CD
Jl
03
>
d
o
4->
O
CD
CO
CO
CO
o
u
u
I— I
O'
<u
bo
03
4->
d
<d
u
CD
a
x
<u
4->
CD
CD
to
c
CD
£
CD
U
d
CO
tti
CD
£
Jh
o
CD
C
03
t
03
>>
TJ
CD
d
u
CO
O
O
o
■M
CD
oJ
CO
CD
d
CD
CO
CD
-d
H
T3
65
Comparing the spectra given in Figures 25 and 26 with Figure 28
and Table 7, we note that the two strongest fluorescence lines are the a + a
transitions with the corresponding b a transitions among the strongest in
the O spectra. As mentioned in Section D. 2. , this is the behavior predicted
from Table 4 for the transitions to a T state having a large *J | = 9/2
7 7 z m
composition where the A_^ parameter is dominant. Although this is the
only case in Table 4 that could clearly produce one strong line while dis-
criminating against the remaining lines in n polarization, other possibilities
may exist. It is sufficient to recognize that Table 4 predicts the possibility.
It is interesting in this regard to compare YVO^ to CaWO^. The
Nd site symmetry in the latter is the group. The only difference in the
formal expansion of the crystal-field potential compared to YVO^ is that
the terms having | q I > 0 are, in general, complex [23]. Yet the spectrum
of NdiCaWO^ [24] does not show a large concentration of the available
oscillator strength into one transition as is observed in NdrYVO^. Since
Table 4 is applicable to CaWO except for the rep labels (see Reference 17),
we conclude that something akin to the A (k odd) being competitive with
7 . iZ
A ^ is the cause for this difference between the two hosts. In other words,
the odd -par ity portion of the crystal field is concentrated in the high -order terms
in YVO, while in CaWO., it is either in the low-order terms or distributed
4 4
over all the terms. This difference appears to be reflected in the even-
6 2
parity terms. The ratio B Q/B Q for Nd in CaWC>4 is -0.031 [220 while
in YVO it is 8.6 Q (Note that B - ( r^)A^ , where ( r^) are the usual
4 q q
radial integrals. )
It can be shown that the odd-parity A are quite sensitive to ion
placements and to slight deviations from the normal site symmetry. The
difference between YV04 and other hosts may be a result of this sensitivity.
A case that illustrates the dependence of line strengths on the ion place-
8 6 3 +
ments is the behavior of the S'* I transitions of Gd in BaF [25]. These
Lr
66
transitions are magnetic dipole forbidden but are observed in CaF and
SrF as electric dipole transitions via intermediate coupling [26, 27] and
Z 5 7
the A and A parameters. However, they are absent in BaF at low
3+ ^ ^
Gd concentrations, say^ 0. 1 atomic %. As the concentration is increased,
these transitions suddenly turn on at around 0. 2 atomic % and grow very
rapidly. It was determined that this effect was due to the minute change in
3 1
the lattice parameter introduced by substituting the smaller Gd ion (1.02 A)
2 t
for the Ba ion (1.351). In simple terms, this small change in lattice
dimension produced large changes in the above odd parity terms while the
crystal field splittings changed by <1 cm
There is the question of the missing transitions in Figure 28 that
are allowed by the selection rules. This is most evident in transitions with
4
the c and d levels of the I . manifold and with the d and e levels of the
4 ^
I nj2 manifold. Regarding the former and referring to the low tempera-
ture a spectrum in Figure 25, we see that one of the transitions a -» d or
b -♦ d is missing. Since d is a T as are levels a and b, then from Table 4
7 *
assuming A ^ is dominant we see that unless d has some 1 = 9/2
content, it will not be observed in it polarization. It will be observed in
a polarization via its |j^j = 7/2 content but only to a T^ level. Thus, we
are led to identify the observed transition to be a d. This choice is
contrary to that reported in previous studies [9, 10, 12], but is consistent
4 2
with the absence of the *g/2^ ^1/2 ^ransi^lon*
There is no ambiguity in the assignment of the transitions with
4 2
level I because of the observed transition to the level. The
corresponding line in the Figure-25 a spectrum is the b c transition.
This line, being anomolously wider than either of its neighbors, is suffi-
cient to include a weak a c line in the high wavelength wing. Regarding
this width, there have been two Raman modes in YVO^ reported at 157
and 162 cm * of symmetry B (T ) and E (T_), respectively [28], Both of
lei e 5
67
these modes will couple the 1^ and 1^ states [27] (i. e. , levels c and a)
which suggests that level c is broadened by a strong coupling with the lattice.
4 4
The distribution of oscillator strength in the *11/2 trans*"
tions can be understood by extending Table 4 to include J = +1 1/2. Again,
7 z
if the A ^ parameter dominates the induced dipole moment, then only those
r? levels containing a large admixture of = +7/2, +9/2 and those T ^ levels
containing J = +11/2 will have strong transitions terminating on them.
Saying it another way, those levels having little or none of these J values
4 Z
will have little or no coupling with the F . levels,
4 3 * 2
Regarding the absence of the *9/2^"^* ^1/2 transition, ^ ls
reasonable to assign this to the Boltzmann depopulation of the level e plus
an oscillator strength smaller than the corresponding TT transition.
F. Summary
A reasonably unambiguous determination of the positions of the
4 4 4 4
crystal-field levels of the I9/2’ ^ll/Z9 ^3/2' anc* ^1/2 ^ manifolds of
Nd in YVO^ has been made. The proposed energy level scheme is consis-
tent with the observed polarized spectra. The ground state was determined
to be P7( JJL = +3/2) with a J composition of +1 /2, ±7/2, and +9/2 consistent
with previous studies [12, 20]. An analysis of the implications of the Judd-
Ofelt theory [l 3]regarding the transitions between the crystal-field levels
showed that (1) lines can be missing, and (2) it is possible for a tt transi-
tion to have a relatively high oscillator strength in the spectra of Nd in
or a similar symmetry. It was found that the unusual disparity of strengths
of the strong lines relative to the remaining lines, particularly in the TT
7
spectra, could be qualitatively understood if the contribution of the A
+0
parameter to the induced electric dipole moments is significantly greater
than the contributions from the A ^ (k = 3,5,7) parameters. This suggests
that it would be of value to compare detailed crystal-field and line-strength
calculations performed for various host crystals in order that the difference
between YVO^ and other laser hosts can be more clearly understood.
68
IV. SURVEY OF Nd: YVO . ABSORPTION
4
The Cary 17 Dual Beam Spectrophotometer was used to measure the
absorption spectrum of Nd:YVO^ from 0. 38 to 2. 0 p m. The lower limit
was determined by the material's uv absorption edge and the upper limit by
the instrument. Several bands of interest were probed. There are at ~ 1.6,
0. 89, 0. 81 , 0. 75, 0. 59, 0. 53, 0. 48 and 0. 44 pm. The observed lines are
tabulated in Table 8. Copies of the data for these bands obtained with a 1%
Nd:YVO^ sample, 4 mm thick, are included for both a and n polarized light
in Figures 29-44.
The most intense absorption line in the - 0.81 pm band is the
0. 8080 pm line for rr polarized light. Its absorptivity is ~ 8. 6 cm The
0.8078 pm line fora polarized light is also strong having an absorptivity
of ~ 4. 2 cm \ Since these lines are part of a strong pump band, it is
therefore recommended that diodes for laser excitation of Nd in YVO, be
4
tuned for 0. 8080 pm and have sufficient line width to also pump at 0. 8078
pm.
The "yellow" absorption band (at ~ 0. 59 pm) is the most intense
absorption band in the NdrYVO^ spectrum. In tt polarized light the line
at 0. 5940 pm has an absorptivity of ~ 1 5. 3 cm There are several other
very intense lines in this group for both a and rr polarized light. Since dye
lasers in this spectral region are available^ laser excitation using these
lines should be possible.
It should be noted that pumping at the most intense absorption wave-
lengths will result in very high thermal loading to a very small volume of
material. This may result in pump induced catastrophic or non-catas-
trophic material failure. In addition, the pumped volume will be restricted
to the region near the entrance surface and so any lasing which occurs
will be similar to that produced by a surface laser. It is suggested there-
fore that dye laser pumping studies be conducted for a variety of wave-
lengths in the nyellowM band so that the optimum combination of absorptivity.
69
excited lasing volume and thermal loading can be found.
In the cw laser experiments we pumped with the 0. 5145 gm laser
line from an argon ion laser. The nearest absorption line for this pump
source is the 0. 5143 |jm line fora polarized light. Even though it pre-
sents a good wavelength match to the pump line, this line is relatively
weak. Its absorptivity is only - 0. 4 cm The absorption at 0. 5145 pm
for TT polarized light is in the edge of the 0. 5118 |jm line and has an
absorptivity of ^ 0.4 cm
The tt and a absorption spectra of a 3%, 8. 5 mm thick, Nd: YVO^
crystal in the 0. 89 |jm band are included in this report. These should
be compared with the same spectra for the 1% Nd:YVO^ sample. It is
clear that the line shapes and position do not change with concentration.
In addition, for all absorption wavelengths in the 0. 7-0. 95 |jm range, the
absorption strengths were found to scale with the Nd concentration.
As a result of our spectral measurements at 1. 6 urn we have
shown the Cary 17 Spectrophotometer to be partially linearly polarized.
In particular, it appears to be partially polarized in the horizontal plane.
Its optical system is essentially the same as that of the Cary 14 and so
we suggest that data taken with such instruments be scrutinized for this
effect. To insure the linear polarization necessary for separation of the
a and TT spectra of the shorter wavelength absorption bands, calcite -Gian
polarizers were inserted in both the sample and reference beams of the
instr ument .
70
Table 8
Absorption Lines of Nd:YVO at Room Temperature
Nominally 1% Nd
Classification
13/2
15/2
a Polarization
tt Polarization
X (|jm)
a (cm'* )
X (pm)
a(cm'*)
2. 564
2. 535
2. 538
2. 518
strongest
2. 480
2.483
2.455
2. 455
2.408
strongest
2.408
1. 766
0. 018
1. 766
0. 081
1. 747
0. 064
1. 748
0. 012
0. 738
0. 055
1.713
0. 035
1. 713
0. 166
1. 705
0. 035
1. 706
0. 129
1.691
0. 042
0. 667
0. 023
1.675
0. 035
1. 638
0. 046
1. 6 50
0. 11 5
1. 627
0. 037
1. 627
0. 075
1. 598
0. 055
1. 598
0. 449
1. 583
0. 216
1. 583
0. 081
1.302
0. 063
1. 288
0. 046
0. 9138
0. 253
0. 9132
0. 293
0. 9030
0. 127
0. 901
0. 144
0.8913
0. 753
0. 8877
0. 805
0. 8868
0. 886
0. 8794
4. 899
0. 8785
1. 500
0.8675
0. 293
0. 8687
0. 196
0. 8577
0. 219
0. 8350
0. 661
0. 8351
0. 288
0.8158
2. 645
0. 8180
1. 323
0. 8078
4. 169
0. 8080
8. 596
0. 7883
0. 270
0. 790
0. 230
0. 7754
0. 230
0. 7745
0. 633
0. 7666
0. 587
0. 7562
3. 726
0. 7530
4. 255
0. 7505
4. 428
0. 7435
1. 783
0. 7440
2. 3 58
0. 6347
0. 075
0. 6302
0. 029
0. 6308
0. 144
0. 6260
0. 173
0. 6262
0. 058
0. 610
0. 219
0. 6101
2. 3 58
0. 6045
1. 610
0. 6048
1. 541
0.6015
3. 996
0. 6010
2. 128
0. 5982
14. 346
71
Table 8. (Continued)
Classification
a Polarization tt Polarization
X (tim)
a(cm-1)
X((jm)
a (cm-1
0. 5981
6. 986
0. 5957
14. 470
0. 5957
7. 744
0. 5940
15. 266
0. 5938
8.424
0. 5890
5. 520
0. 5892
10. 810
0. 5866
5. 923
0. 5870
4. 313
0. 5832
6. 383
0. 5853
3. 565
0. 5795
1. 265
0. 5795
5. 750
0. 5770
0. 661
0. 5755
0. 431
0. 573
0. 316
0. 5735
0. 259
0. 570
0. 259
0. 5715
0. 173
0. 566
0. 219
0. 5680
0. 144
0. 563
0. 219
0. 5661
0. 127
0. 560
0. 219
0. 5630
0. 115
0. 5600
0. 11 5
0. 5555
0. 086
0. 5525
0. 058
0. 5510
0. 058
0. 5452
0. 086
0. 5523
0. 173
0. 5437
0. 104
0. 5433
0. 863
0. 5408
0. 288
0. 5408
2. 444
0. 5382
0. 977
0. 5382
0. 058
0. 5362
1. 484
0. 5358
0. 058
0. 5380
0. 391
0. 5333
0. 661
0. 5362
0. 903
0. 5301
1. 524
0. 5325
3. 939
0. 5285
0. 288
0. 5285
1. 926
0. 5245
0. 604
0. 5250
0. 010
0. 5222
1. 26 5
0. 5223
1. 236
0. 5200
2. 128
0. 5202
0. 288
0. 5170
1. 254
0. 5178
0. 201
0. 5118
1. 495
0. 516
0. 173
0. 5095
0. 920
0. 5143
0.403
0. 5070
0. 345
0. 511
0. 288
0. 5050
0. 086
0. 5097
0. 431
0. 5070
0. 288
0.4912
0. 029*
0.4970
0. 029*
0. 4855
0. 086
0.4912
0. 069
0.4838
0. 219
0. 4885
0. 012
0.4820
0. 230
0.4850
0. 259
0.4790
0. 184
0.4819
0. 805
0.4760
0. 288
0.4790
0. 144
72
Table 8. (Continued)
Classification
g Polarization tt Polarization
X (um)
a(cm'*)
MUm)
a (cm ~
0.4735
0. 115
0.4759
2. 156
0.4710
0. 132
0.4713
0. 633
0.468
0. 173
0.4670
0. 328
0. 465
0. 219
0.4625
0. 230
0. 4632
0. 276
0.4420
0. 098
0.4372
0. 173
0.4371
1. 064
0.4360
0. 086
0.4340
0. 058
0. 4340
0. 943
0.4250
0. 069
0.4235
0. 058
0. 4235
0. 1 55
* The absorptivity for all
for the broad color center
o
lines at wavelengths <5000A has been corrected
band absorption evident in Figures 43 and 44.
73
?.483 — i
Figure 29. 2. 5 pm Band of Nd:YVO^ Absorption Spectra at Room
Temperature. Taken with Beckman DK-2A Spectro-
photometer. 1% Nd. 4 mm thick.
74
OPTICAL DENSITY
Figure 30. 1.6 |_im Band of NdiYVO^ Absorption Spectra at Room
Temperature. Partially polarized light propagating along
the a -axis. Mostly IT spectrum. 1% Nd. 4 mm thick.
75
A1ISN30 IVDIldO
<1>
>
u
3
o
u
<D
o
*1
u w
5 c
a S
a .
AS
u
aJ
(u j: a
a
c c m #
c £ — o
« g k 2
H ^ 0) i
£
o
o
«
at EF
I— <
aJ
4-> •
u CO
a ’*
CO f
C u
.2 M
a §
3 .— t
O aJ
J3 c
5 .2
■*f aJ
O w>
S rt
r a
>< o
•* ^
5 &
S E
§ E
-*-* _H
o ^
-3
a, .
O X
^ Z
-M
U
d) o
a ^
co
u
at
u
a>
-C
-3
u
rt
0)
.3
X
aJ
i
cti
bo
c
o
3
o
*4->
at
bJD
aJ
a
o
XI
3
0)
3
*-<
4->
rt
4->
rt
•
CQ
u
u
X
<D
4->
0)
£
a
03
03
3
N
d.
O
*-t
xO
D
£
J2
•
co
0)
o
X
a,
a>
3
bJD
£
76
optical density
Figure 32. 0. 89 pm Band of Nd:YVO^ Absorption Spectra at Room
Temperature, o spectrum. 1% Nd. 4 mm thick.
77
OPTICAL DENSITY
Figure 33. 0. 89 |Jirn Band of NdrYVO^ Absorption Spectra at Room
Temperature, aspectrum. 3% Nd. 8. 5 mm thick.
78
OPTICAL DENSITY
Figure 34.
0. 89 Mm Band of Nd:YVO^ Absorption Spectra at Room
Temperature, n spectrum. 1% Nd. 4 mm thick.
79
OPTICAL DENSITY
Figure 35.
0. 89 nm Band of NdrYVO^ Absorption Spectra at Room
Temperature. TT spectrum. 3% Nd. 8. 5 mm thick.
80
0.5
A1ISN3Q IVOIldO
81
Figure 36. 0. 75 and 0. 81 |jm Bands of Nd:YVO^ Absorption Spectra at Room Temperature
a spectrum. 1% Nd* 4 mm thick.
0.8
A1ISN3Q “IVDIldO
82
<u
u
3
4->
rt
Jh
<L>
a
£
H
£
O
o
«
nJ
Jh
4->
U
a 5J
w .c
d ^
.2 g
a g
O ^
CQ
rQ
<;
no
o ^
> -*
>*
• • •
5 £
o -*-»
u
03 <U
TD CL,
C «
rt i—
PQ •=
g
d-
i-H
00
•
o
a?
c
CTj
un
r-
r-
m
<u
O
buo
A1ISN3Q "IVOIldO
83
( V
u
g
rt
u
V
a
E
a>
H
E
o
o
£ u
u £
<u Xi
a „
w £
c g
o M
4-» ^
£X
u .
O TJ
<q s£
O
>
!h
£
TJ
G
cxj
«
£
G
Jh
-4-J
u
a)
a
m
O
■ I
■ i
v
><
£
o
tn
00
CO
<D
G
be
£
i .Orr
84
spectrum. 1% Nd. 4 mm thick.
Figure 40. Yellow Band of NdrYVO^ Absorption Spectra
at Room Temperature. TT spectrum on scale.
1% Nd. 4 mm thick.
85
OPTICAL DENSITY
Figure 41. 0. 53 pm Band of NdrYVO^ Absorption Spectra at Room
Temperature, a spectrum. 1% Nd. 4 mm thick.
86
OPTICAL DENSITY
1.0
0.9
0.8
0.7
0.6h
0.51—
WAVELENGTH (Mm)
Figure 42. 0. 53 pm Band of Nd:YVO^ Absorption Spectra at Room
Temperature, rr spectrum. 1% Nd. 4 mm thick.
87
0.8
£
x
K
O
z
UJ
UJ
>
<
£
A1ISN30 nVDIidO
6
(h
3
rt
CD
a
£
CD
H
£
o
o
rt
rt
4->
u
CD
a
in
c
o
4->
a
u
o
CO
rO
<q
O
>
>-»
ro
£
o
co
TD
C
aJ
PQ
£
n
lT>
O
V
CO
CD
U
3
W)
88
a spectrum. 1% Nd. 4 mm thick.
O O'
-* d
A1ISN3Q IVOIldO
89
Figure 44. \< 0. 5 pm Bands of NdiYVO^ Absorption Spectra at Room Temperature.
rr spectrum. 1% Nd. 4 mm thick.
V. LASER PERFORMANCE TESTING
A. Laser Testing
During the second half of this project, no Nd doped YVO^ laser rods
suitable for testing were delivered by the subcontractor. Those rods that
were delivered were either too short to satisfy the requirements of this
program or too lossy for meaningful laser testing due to the presence of
particulate scattering centers. We were therefore unable to conduct further
pulse pumped Q switched laser testing of Nd:YVO^. In addition, though
Ho, Er, Tm:YVO^ laser rods were delivered, we suspended the planned
effort to study lasing from Ho in YVO^ at the request of USA -ECOM.
We did, however, evaluate the effect of the ignitron in the flashlamp
circuit on the laser’s performance. As shown in Figure 45, the effect of
the ignitron was negligible. We also tested a Nd:YAG rod in pulse pumped
Q switched operation in our laser cavity. This laser rod was obtained on
loan from USA-ECOM to provide data which would demonstrate the quality
of our pump cavity and laser cavity optics. It would also be data which
would permit easy intercomparison of our system with that in use at ECOM.
Figure 46, shows this data obtained from the ECOM Nd:YAG rod and for
comparison. Figure 47 is a reproduction of Figure 19 of the semiannual
report which shows the Q switched performance of Nd:YVO^ laser rod, 3L.
B. Electrooptic Q Switching of the Nd:YVO^ Laser Without an
Intracavity Polarizer
Pulse-pumped Nd:YAG lasers are often Q switched using the Pockels-
cell-Glan- polarizer combination shown in Figure 48(a). Since the host
material is optically isotropic, both the Pockels cell and the Gian polarizer
must be placed in the optical cavity to prevent premature lasing or pre-
pulsing. As indicated in Figure 48(a), when the voltage applied to the Pockels
cell produces a one-quarter wave, X/4, phase retardation at the laser wave-
length, the light reflected from the 100-percent R -cavity mirror enters the
90
0 10 20 30
Ellgl JOULES
Figure 45. Nd:YAG laser output vs. input energy data with and
without the ignitron in the discharge circuit. Current
pulse FWHM is 125 ^ls with a krypton-filled lamp,
1 200 Tor r.
91
INPUT ENERGY, JOULES
Figure 46. Output versus input energies for the 3 x 30 mm Nd:YAG
(ECOM-YAG) using the Xe flashlamp (450 T) and the
125 psec (FWHM) pump pulse duration. Output R =45%.
92
Figure 47. Q-switched laser output vs. input energy from
Nd:YVO^ (rod 3L) using a Pockels cell Q-switch
with no intracavity polarizer. The output reflector
was 3%R at 1.06 ^im. x Pockels cell Q-switched.
O Long pulse lasing with Pockels cell in cavity.
□ Long pulse lasing, empty cavity.
93
100% R
MIRROR
POCKELS
CELL
(a)
-e®
X/4
VOLTAGE
ON
OUTPUT
MIRROR
Laser rod can not
amplify light _L to the
sketch
Laser rod is
birefringent and does
not have parallel faces
Figure 48. Electrooptically Q-switch configurations, (a) A Pockels -cell-
Glan- polarizer combination in a Nd: YAG laser cavity, (b) A
Pockels cell only in a cavity which has a strong polarization-
dependent gain, (c) A Pockels cell in a cavity where the laser
rod is slightly wedged and strongly birefringent.
94
Gian polarizer polarized so that it is not transmitted. The combination of
intracavity Pockels cell and Gian polarizer thus spoils the cavity Q by
isolating the 100-percent R mirror from the optical resonator. When the
voltage applied to the Pockels cell is reduced to zero, there is no phase
retardation, the cavity Q is restored, and an intense Q-switched pulse of
laser light is^produced.
A significant simplification of the Q-switched laser could be achieved
if the optical resonator did not require an intracavity polarizer for prepulse
free operation. This was attempted by McClung and Hellwarth C 2 9J for a
ruby laser and by Weber et al. [3 0 3 for a NdtYAlO^ laser. In both of these
cases, the authors attempted to use the anisotropic properties of the laser
gain in a crystalline host to prevent premature oscillation in the laser cavity.
Consider a laser cavity where the active medium has gain for light
polarized in one direction only. After two complete round trips through this
cavity with no intracavity polarizer and X/4 voltage applied to the Pockels
cell (see Figure 48(a) ), a condition for oscillation can be reached. If this
condition
(Rj)2 exp (2 Ph£) = 1
is met, there will be prepulsing. Here is the output mirror reflectivity,
V
B is the net laser gain (R is assumed to be zero for light polarized normal
H H
to the figure in Figure 48(b) ), and H is the length of pumped medium. At
threshold for lasing with no voltage applied to the Pockels cell, the oscilla-
tion condition is
Rj exp (2 3Th^ ) = 1.
Thus 3 = 2 3 and so no holdoff is possible at pump levels which produce
H Th
a gain equal to twice that at threshold. This is in agreement with the results
in [29] and [30], R ecently, however, prepulse-free electrooptic Q switching
of Nd:YVO^ with no intracavity polarizer was reported at pump levels as
95
high as eight times threshold [7 ] (see Figure 47). Since the anisotropic
gain of the laser crystal cannot account for such excellent performance, the
explanation must lie elsewhere.
The NdrYVO^ laser rod studied in [ 7 ] was inadvertently polished so
-3
that its flat end faces were wedged with respect to each other by ~6 x 1 0
rad. This, coupled with the very strong birefringence of the YVO^ host
3
(n = 1. 958 and n = 2. 168 at 1. 06 ^lm ), permitted the laser rod to act as
o e
its own intracavity polarizer. To understand the importance of the wedge
between the rod faces consider that if the flat, 0. 25-percent R faces had
been parallel, laser action would have occurred between the rod faces at
pump levels ~3. 5 times that required for lasing in the main optical resonator
(100% and 3% R mirrors). Since no oscillations at all were reported with
the optical-cavity mirrors misaligned, the wedge between the rod faces is
sufficient to prevent unwanted lasing, and therefore is a critical factor in
determining the properties of the Nd:YVO^ laser.
Suppose, for simplicity, that the face of the rod nearest the 100-
percent R mirror was normal to the rod axis and that the other face was
.3
misaligned by the wedge angle (6 x 10 rad) (see Figure 48(c) ). Since the
cavity mirrors were aligned for best long-pulse lasing, they were oriented
so that light polarized parallel to the direction having maximum gain propa-
gated parallel to the rod axis. In [ 7 ] the rod was nominally cut with its
axis parallel to the crystalline a axis and, since maximum gain in Nd:YVO^
occurs for light polarized parallel to the optic axis, light which propagated
as an extraordinary ray in the laser rod was that which could oscillate.
When the Pockels cell voltage was on, light reflected from the 100-
percent mirror entered the rod as an ordinary ray. As it left the Nd:YVO^
rod it was not refracted far enough to be normal to the output mirror, and
so no premature oscillations were possible. For the wedge angle of
6x10 rad of the NdrYVO. laser rod in [7 J the extraordinary ray would
4
96
have been refracted by 14x10 rad and the ordinary ray by 11.8x10
-3
rad. The resulting 2. 2 x 10 rad (0. 126°) misalignment would have been
sufficient to prevent prepulsing, and thus explains the excellent results
reported.
The observation that a suitably birefringent laser crystal can be
prepared so that is can act as its own intracavity polarizer for electrooptic
Q switching can lead to great simplification in Q-switched laser systems.
This was demonstrated in the present correspondence for NdrYVO^ and can
be adapted to the design of other birefringent laser crystals.
97
VI. CONCLUSIONS AND RECOMMENDATIONS
FOR FUTURE EFFORTS
Using a laser rod designed according to the results of our spectro-
scopic studies we have demonstrated excellent pulse-pumped Q-switched
laser performance from Nd:YVO^. Due to the inherent birefringence of the
host material, a wedged laser rod made possible pre-pulse free Q switching
with no intracavity polarizer. This simplifies the design of laser systems
using Nd:YVO and can significantly reduce their cost.
3 +
Because of the high cross section for stimulated emission of Nd
in YVO^, we foresee major uses of this material in pulse pumped and CW
1. 06 fim laser systems where the available pump power is limited. In
addition, where simplicity is essential, Q-switched lasers at 1.06 \j. m can
benefit from the use of Nd:YVC> laser rods.
4
The major problems to be solved in the use of NdrYVO^ in laser
systems is reliable crystal growth and fabrication techniques. Inclusions
lying along cleavage planes cause cracking during cool-down and sometimes
during rod fabrication. These inclusions also cause scattering and in this
project resulted in Ho, Er, Tm:YVO^ crystals that were too lossy to lase.
We recommend future efforts to 1) develop crystal growth and fabri-
cation techniques, 2) study CW lasing of NdrYVO^ and 3) study Ho, Er, TmrYVO^
lasing near 2 Hm.
98
REFERENCES
1. J. R. O'Connor, Appl. Phys, Lett, 9, 407 (1966).
2. R. J. Pressley, "Improving Solid-State Lasers," RCA Laboratories, 1969.
3. H. G. McKnight and L. R Rothrock, "Research and Development Work
for the Growth of Single Crystal Yttrium Orthovanadate," Technical
Report ECOM0022F, U. S. Army ECOM Contract DA A B07 -72-C-0022,
April 1973.
4. H. M. Dess and S. R. Bolin, paper presented at Electronics Materials
Conference, Boston, Mass., Aug. 1966.
5. H. M. Dess and S. R. Bolin, Trans. Metal. Soc. A I ME 239, 359 (1967).
6. L.G. DeShazer, M. Bass, U. Ranon, J.K. Guha, and E. D. Reed,
VIII Intern. Quant. Elect. Conf. , June 1974 (IEEE, New York), pp. 708.
7. M. Bass, L.G. DeShazer, and U. Ranon, Technical Report ECOM-74-
0104-1 (October, 1974), U. S. Army ECOM, Fort Monmouth, New Jersey.
8. L.G. DeShazer, A.W. Tucker, M. Birnbaum, and C. L. Fincher, 1975
IEEE/OSA Conference on Laser Engineering and Applications, Digest of
Technical Papers (IEEE, New York, 1975), Post Deadline Paper No. 4. 10.
9. Kh. S. Bagdasarov, G.A. Bogomolova, A. A. Kaminskii, and V. I. Popov,
Dokl. Akad. Navk SSSR 180, 1347 (1968) [Sov. Phys. Dokl. J_3, 516
(1968)].
10. A. A. Kaminskii, G.A. Bogmolova, and L. Li, Izvestiya Akad. Nauk.
SSSR, Neorganicheskie Materialy 5, 673 (1969).
11. R.J. Pressley, P. V. Goedertier, and H. Weakliem, rMBT 70 Laser
Materials Research and Exploratory Development,11 RCA Lab, Princeton,
New Jersey, Report (October 1969).
12. N. Karayianis, C.A. Morrison, and D. E. Wortman, Harry Diamond
Laboratories, Washington, D. C. (unpublished).
13. B. R . Judd, Phys. Rev. 1 27, 750 (1962);
G.S. Ofelt, J. Chem. Phys. 37, 51 1 (1962).
99
14. L. R Rothrock and R. E. Wilder (work reported in References 3 and 8).
15. R.W.G. Wyckoff, Crystal Structures, (Inter science. New York, 1 963).
16. K.H. Hellwege, Ann. Physik 4, 95 (1948).
17. G. F. Roster, J. O. Dimmock, R.G. Wheeler, and H. Statz, Properties
of the Thirty-Two Point Groups (M. I. T. Press, Cambridge, Mass.,
1963).
18. J .A. Koningstein and J. E. Geusic, Phys. Rev. 1 36, A711 (1964).
19. K. Rajnak, J. Chem. Phys. 43, 847 (1965).
20. U. Ranon, Phys. Letters 28A, 228 (1968).
21. C. Brecher, H. Samelson, A. Lempickl, R. Riley, and T. Peters, Phys.
Rev. 155, 178 (1967).
22. N. Karayianis and R.T. Farrar, J. Chem. Phys. 53, 3436 (1970).
23. J. L. Prather, ’’Atomic Energy Levels in Crystals,” NBS Monograph 19
(Superintendent of Documents, Washington, D. C. , 1961).
24. J. G. Gualtieri and T.R. Au Coin, J. Chem Phys. 45, 4348 (1966).
25. R.C. Rang and P. P. Yaney, Bull. Am. Phys. Soc. 19, 818 (1974);
R.C. Rang, M. S. Thesis (University of Dayton, 1974) (unpublished).
26. R. E. Ziegler, P. P. Yaney, and J.A. Detrio, Technical Report AFML-
TR-69-230 (December 1969) Air Force Materials Laboratory, Wright-
Patterson AF Base.
27. J.A. Detrio, M. W. Ferralli, P. P. Yaney, D. M. Ware, and V. L. Dolan,
J. Chem. Phys. 53_, 4372 (1970).
28. R.T. Harley, W. Hayes, and S.R.P. Smith, Solid State Commun. 9, 515
(1971).
29. F.J. McClung and R . W. Hellwarth, J. Appl. Phys. 33, 828, (Mar. 1962).
30. M.J. Weber, M. Bass, K. Andringa, R.R. Monchamp and E. Comperchio,
Appl. Phys. Lett. 1 5, 342 (Nov. 1969).
100
APPENDIX A
IF-F.E JOURNAL OF QUANTUM ELECTRONICS. VOL. QE-11, NO. 3, MARCH 197S 97
Analysis of Laser Emission in Ho3+ -Doped Materials
JOHN A. CAIRO and LARRY G. DeSHAZER
Abstract -The oscillator strengths, transition rates, and branching
ratios associated with known and potential laser transitions in trivalent
holmium are discussed. Recently reported laser transitions between
the 5Fj and 5lj energy manifolds are found to have an abnormally
low average oscillator strength, and an alternative explanation of the
laser observation is suggested.
1. Introduction
INDUCED electric dipole oscillator strengths and transition
rates between energy levels of rare-earth ions in crystals
can be calculated using a theory developed by Judd [1] and
Ofelt [2]. In recent years the Judd-Ofelt (JO) model has
been used in the analysis of rare-earth laser materials [3] -[6] ,
and its use led to the discovery of the four-level Tm:Cr:YAl 03
laser at 2.35 pm (7]. This paper reports the results of an
application of the JO model to known and potential laset
transitions in trivalent holmium.
spectrum. Equation (1) may then be used to calculate
oscillator strengths in the fluorescence spectrum. It should
be noted that (I) gives the oscillator strength summed over
the Stark split components of the final energy manifold f
and averaged over the components of the initial energy
manifold / .
The cross section for stimulated emission integrated over all
Stark split components of the upper and lower /-manifolds
is given by [5]
<3>
where i> = 1/X. Using (2) this reduces to
f we3
J o(v)dv=—jfjj' (4)
J-+J }TlC
II. Theoretical Background
According to the JO model, the oscillator strength / and
spontaneous emission rate A of crystal field induced electric
dipole fluorescence in a rare-earth ion are given by [8]
fjj '
8 n7mc
3h\(2J + 1)
V +2)*1
9 n J
and
• 21 n,\Wn[S. L\J\\Ut\\Afn [S\L,}/)\1
f-2.4,6
8/r’eV
Ajr = fjj
(i)
(2)
where X is the mean wavelength associated with the transition,
n is the crystal’s index of refraction at the mean wavelength,
and 2/+ 1 is the number of levels in the upper /-manifold.
In (l) the <|| are doubly reduced matrix elements of unit
tensor operators calculated in the intermediate coupling
approximation [ 1 ) , and the £2, are three empirical parameters.
These empirical parameters can generally be determined for
each ion-host combination by a least squares fit between
calculated and measured oscillator strengths in the absorption
Manuscript received June 7, 1974; revised October 16, 1974. This
work was supported in part by the Air Force Systems Command, Air
Force Avionics Laboratory, Wright-Patterson Air Force Base, Ohio,
and in part by the U.S. Army ECOM under Contract DAAB07-74-C-
0104.
J. A. Caird is with the Department of Physics, University of Southern
California, Los Angeles, Calif. 90007, and with Hughes Aircraft Com
pany, Culver City, Calif. 90230.
L. G. DeShazer is with the Departments of Electrical Engineering,
Physics, and Materials Science, University of Southern California, Los
Angeles, Calil'. 90007.
which shows the significance of the oscillator slrength in laser
analysis. Without detailed line shape information (4) cannot
be used to predict peak cross sections, but it is evident that
transitions with larger oscillator strengths will generally have
larger peak cross sections.
In the 4 fn ground configuration of rare-earth ions, electric
dipole transitions become allowed only when the ion feels a
perturbing field lacking inversion symmetry so that parity
restrictions are broken. Magnetic dipole (md) transitions
between levels with | A / < 1 1 are always allowed, however, and
can on occasion be comparable in strength to the crystal field
induced electric dipole (ed) transitions A definition of the md
oscillator slrength consistent with (2) and (4) is given by [9]
h
fn • (mtl)^-T —
6X(2/ + 1 )mc
• n |<4/n [5, L\J\\L + 2511 4 /" [S’, /.') /> I5 (5)
where the matrix elements of L + 25 can be evaluated using
standard formulas given elsewhere [10] .
A computer program was developed capable of calculating
the unit tensor operator matrix elements in (1) between all
/-manifolds in all rare-earth ions using previously published
intermediate coupling wave functions [6], The program was
then used to calculate ed and md oscillator strengths, transi-
tion rates, and radiative branching ratios for a large number of
ton-host eombmalions for which intensity parameters £2, had
been published. The branching ratio for a transition / -►/is
simply given by
Pi
i
X) ^ (i —»/')
(6)
Copyrighl 1975, by th* institute of Electrical and tlectionics Engineers, Inc.
PRINTED IN THE U.S.A. Annib No. 503QI-001
98
IEEE JOURNAL OF QUANTUM ELECTRONICS, MARCH 197S
where the sum in the denominator is over all states of lower
energy. In these calculations emphasis was placed on transi-
tions for which laser action and/or fluorescence has been
reported, and those for which the radiative transition rates
were large enough to exceed multiphonon emission rates. It
was found that the vast majority of laser lines had oscillator
strengths greater than I0~6. In a few notable exceptions
although the oscillator strengths were somewhat lower than
10 6, the branching ratio associated with the laser transition
was particularly high, and the upper level lifetime was long,
allowing for the buildup of a large inversion density. The
only other exceptions were found in cases where a trivalent
rare-earth ion was placed on a divalent ion site in the laser
material.
III. Application to Holmium
Laser emission from Ho3* ions has been observed in many
hosts on transitions between the 5I7 and 5I8 energy mani-
folds at wavelengths near 2 jrm. Laser action in the green
near 0.55 fjm has also been reported on [5S2, 5F4] -*SI8
transitions in CaF7 [II] and BaY7F8 [12]. Recently
stimulated emission in Ho:BaY7F8 near 2.4 jxm has been
attributed to transitions within the SF5 “*SI5 group [13].
Fluorescence originating in the 5I6 state has also been ob-
served in a number of crystals [9] , [13] , [14] , but emission
from other levels is often quenched by nonradiative processes.
The intensity parameters in the JO model for Ho:YA103
were determined by Weber et al [9] to be 1.82, 2.38, and
L53 (X \0~7° cm7) for H4, and f26, respectively. The
electric dipole oscillator strengths and transition rates for
fluorescence from low lying energy levels of Ho*3 in YA103
can be easily derived from data given in [9] along with the
indices of refraction of YA103 given in [15]. These are
tabulated for transitions from the 5I7, SI6, 5FS, and [sS7,
5F4] states in Table I. Calculation of the md contributions
in Table I required the use of intermediate coupling wave
functions for Ho3* which were obtained from the authors of
[9] . The branching ratios are in close agreement with those
reported in [9], with the exception of the 5FS “>SI4 transi-
tion for which the ratio is found to be much smaller than
previously reported.
Examination of Table I shows that the laser transitions
Sl7 “>5I« and [5S7, 5F4]-*sI8 have oscillator strengths
greater than I0-6, consistent with most other rare-earth lasers
[6] . It is interesting to note that 5F4 -»SI8 transitions may
be more important than 5S7 -*5I8 transitions for green laser
emission, and therefore the relative positions of the sS7 and
*F4 energy levels may have a large effect on laser performance.
The 5S7-*5I7 transition at 0.76 Jim also appears to have
sufficient strength to support laser oscillation, but the SI7
lifetime is longer than the 5S7 , so that quenching of the 5I7
level may be required. Conditions also appear to be favorable
for laser emission on both the 5I6 5I7 line at 2.9 jxm and the
sl6 *>SI8 line at 1.2 Jim, particularly since the fluorescence
lifetime of the 5I6 level is reasonably long. At least one
attempt to achieve laser action on the 516 ->SI7 transition,
TABLE I
Oscillator Strengths, Transition Rates, and Branching Ratios
for Ho YAJO3
U»tl
L
Uv.1
A««raga
Rm-lp
W. <1
Cac Halo t St
rangtM»10S
_ -i
2«4.a*tv«
ftrar tJurtj
Ratio
altctnc
dipota
magnate
dipola
• U« if il
<*») •
maj -«t>c
i v-.I*
\
\
11)0
It?
O.tt
tl)
47
I.S2I
\
’*7
1)00
0.41
0. 74
IO
22
.14
\
1)40
>.)4
-
210
. 14
\
1)00
. oct
. 001
. 1 1
. M
t . ooot
\
4)10
.It
.02
12
1.0
.0012
\
1)1»
1. )t
.01
ItO
1. )
.0»1
\
10410
a. it
-
770
-
.IB
\
1)4)0
). 4t
1270
~
.7#
\
>740
0.04
-
0. t
-
. C002
\
M10
1.00
-
41
-
.0:4
\
tow
».)0
-
42
-
. 01)
\
wvo
1.12
240
-
. 0)4
\
mw
).)•
14)0
-
. It
\
It 1 40
2.42
-
11)0
-
■ S4 1
\
low
0.41
0. 1)
9.4
4.1
. 002
\
))t0
0.))
. 001
1)
.07
.CO)
\
7440
1.7)
. oo)
240
.41
. 0)1
\
10000
I. W
-
4)0
-
.04 4
\
moo
1.41
-
710
.C)l
\
10)40
7.01
—
4140
—
.10
however, was unsuccessful [13] even though the *I7 lifetime
was significantly shortened by addition of Pr3* and Eu3* ions.
The oscillator strength of the 5FS -*SI5 emission is seen to
be very small in comparison to other rare-earth lasers, and to
the other holmium lasers in particular. The transition does
not have mitigating advantages either. The upper level life-
time is not long, the lower level lifetime is long (predicted),
and the radiative branching ratio is very small. It would seem
that stimulated emission from the 5F5 level and terminating
on the 5I6, 5I7, and 5I8 levels should have been observed
first. One might ask if the intensity parameters for BaY7F8
could be much larger than in YAJ03. The predicted radia-
tive lifetimes for the SI6 and 5I7 levels in YA103 are 3 and 6
ms, respectively, while the measured lifetimes in BaY7F8 are
5 and 16 ms, respectively [13], indicating that the intensity
parameters for BaY2F8 must in fact be smaller than for
YA103. In addition, examination of the unit tensor operator
matrix elements involved [9] shows that changing the rela-
tive magnitudes of the three intensity parameters could not
change the branching ratio appreciably. It could be that the
stimulated emission cross section for this transition is sharply
peaked, while the integral (4) remains small. However, laser
oscillation was reported on three different lines 2.362 /im,
2.375 jim, and 2.377 jim.
Our own work on the Tm:YAJ03 laser at 2.3 jxm [16]
leads us to suggest that the emission seen by Johnson and
Guggenheim may have been due to impurity thulium ions in
their laser rod. It should be noted at the outset, however, that
the authors of [13] indicate that the Ho:BaY7F8 laser
CAtRD AND DE SHAZER ANALYSIS OK LASER EMISSION
99
26 _
24 -
Fig. 1. Energy levels of Ho3*, Tm3*, and Er3* indicating transitions
in the 2.3-2.4-*un region.
material was of very high purity and that no fluorescence
typical of thulium was observed [17].
Fig. I compares the energy levels and laser transitions of
Ho34 and Tm3+ in question. Er3* is included for later discus-
sion. The laser wavelengths which have been observed in
Tm.YA103 on 3F4 ->,3H5 transitions are 2.348 /xm, 2.349 Atm
[18], and 2.272 /im [19], while the range of possible wave-
lengths in this group extends from 2.13 to 2.61 /xm based on
the energy levels given in [18]. Also indicated in Fig. 1 are
possible mechanisms for nonradiative transfer from Ho3* to
Tm3*. This excitation scheme is consistent with the observa-
tions of Johnson and Guggenheim since they indicate laser
oscillation is initiated by pumping the 5F5 level of Ho3*.
Nonradiative transfer from Ho3* SI4 to Tm3* 3F4 is also pos
sible, but direct excitation of the 5I4 level of Ho3* is ineffec-
tive because the absorption of the 5I4 level is extremely weak.
Cessation of oscillation early in the pump pulse attributed to
terminal state population buildup was observed [13] and may
be consistent with the Tm3* 3H5 level lifetime in BaY2F8
(though this is not observed in YA103). The normal thulium
aH4 -►3H6 fluorescence at 1.9 /mi would be quenched by
nonradiative energy transfer from the Tm3* 3H4 to the
Ho3* 5I7 level, and would therefore be difficult to observe.
Fluorescence from the 3F4 -+3H6 transition at 0.8 /im might
be expected, however, if thulium were responsible for the
laser action. One might ask why the Tm3* transition is not
observed to lase in a large number of crystals in which Ho3*,
Er3*, and Tm3* are doped together to sensitize the 2-/xm
transition in holmium. Reference to Fig. 1 shows that the
presence of erbium in these cases would quench fluorescence
from the 3F4 level of Tm3* by nonradiative energy transfer
to the 4I9^ level of Er3*.
IV. Summary and Conclusion
An analysis of holmium laser transitions has been performed
based primarily on the Judd-Ofclt theory of crystal field
induced electric dipole oscillator strengths. The JO model
appears to be useful in describing some of the actual and
potential laser properties of Ho3*, except in the case of the
reported 5FS ~>5IS transitions. The observation of laser
emission from holmium on these lines would not indicate a
b;cakdown of the JO model, but would show that the
applicability of the JO model to laser analysis is limited in its
present form. If, however, the observed emission could be
attributed to thulium impurities in the crystal, the JO model
will have been a useful tool in the identification and prediction
of the laser transitions.
References
[1] B. R. Judd, “Optical absorption intensities of rare-earth ions,”
Phys Rev.tv ol. 127,pp. 750-761,1962.
[21 G. S. Ofelt, “Intensities of crystal spectra of rare-earth ions,”
J Chem. Phys., vol. 37, pp. 51 1-520, 1962.
[3] W. F. Krupke, “Radiative transition probabilities within the 4/3
ground configuration of Nd:YAG,” IEEE J. Quantum Electron. ,
vol QE-7, pp 153-159, Apr. 1971
[4] , “Assessment of a promethium YAG laser,” IEEE J. Quan*
turn Electron. (Corresp.), vol. QE-8, pp. 725-726, Aug. 1972.
[5] , “Induced-emission cross sections in neodymium laser
glasses,” IEEE J. Quantum Electron., vol. QE-10, pp. 450-457,
Apr. 1974
[6] J. A. Caird, “Theoretical analysis of rare-earth laser materials,”
in the Optical Society of America 1974 Spring Meeting Program
(Washington, D.C.), Apr. 21-25, p. 535.
[7] L. M Hobrock et al., “Four-level operation of Tm: Cr: YAIO3
laser at 2.35 jim,” in Dig. Tech. Papers, VIII Int. Quantum
Electronics Conf (Montreal, P. Q., Canada), May 8-U , 1972.
[8] W. F. Krupke, “Optical absorption and fluorescence intensities
in several rare-earth doped Y203 and LaF3 single crystals,”
Phys. Rev., vol. 145, pp. 325-337, May 1966.
[9] M. J. Weber, B. H Matsinger, V. L. Donlan, and G. T. Surratt,
“Optical transition probabilities for trivalent holmium in LaF3
and YA103,”y. Chem. Phys., vol. 57, pp 562-567, July 1972.
[10] B. G. Wyboume, Spectroscopic Properties of Rare Earths . New
York: Interscience, 1965.
[11] Y. K. Voronko, A. A. Kaminsky, V. L. Osiko, and A. M.
Prokhorov, “Stimulated emission of Ho3* in CaF2 at 55 12 A,”
Sov. Phys.- JETP Lett. , vol. 1 , pp. 3-5, Apr. 1965.
[12] L. F. Johnson and H. J. Guggenheim, “lnfrared-pumped visible
laser,” Appl. Phys. Lett,, vol. 1 9, pp 44-47, July 1971.
[13] — , “Electronic- and phonon -terminated laser emission from
Ho3* in BaY2F$,” IEEE J. Quantum Electron., vol. QE-10, pp.
442-449, Apr. 1974.
[14] G. H. Dieke, Spectra and Energy Levels of Rare Earth Ions in
Crystals . New York: Interscience, 1968.
[15] K. W. Martin and L. G. DeShazer, “Indices of refraction of the
biaxial crystal YA103,” Appl. Opt., vol. 12, pp. 941-943, May
1973.
116] J. A. Caird, J. Nella, and L. G. DeShazer, unpublished.
[17] L. F. Johnson, private communication.
[18] L. M. Hobrock, “Spectra of thulium in yttrium orthoaluminate
crystals and its four-level laser operation in the middle infrared,”
Ph.D dissertation, Univ. Southern California, Los Angeles, Jan.
1972.
[19] J. Nella, private communication.
APPENDIX B
Continuous-wave operation of Nd:YV04 at 1.06 and 1.34 ^
A. W. Tucker, M. Birnbaum, and C. L. Fincher
Electronics Research Laboratory ; The Aerospace Corporation. El Segundo, California 9 0245
L. G. DeShazer
Center for Laser Studies, University of Southern California, Los Angeles, California 90007
(Received 7 August 1975)
Continuous-wave plane-polarized outputs of 1 W at 1.06 p. and 0,35 W at 1.34 p. were obtained by end
pumping small samples of Nd : YV04 with an argon-ion laser. Slope efficiencies and material losses were
determined. At 1.06 and L34 p., Nd: YV04 lasers can substantially outperform Nd.YAG lasers. Self- Q-
s witched operation of Nd:YV04 at both wavelengths was obtained by resonator misalignment
PACS numbers: 42.60 G
Lasers of high efficiency and icrw thresholds are re-
quired in diverse applications such as communications,
ranging, and metrology. The Nd : YAG laser has repre-
sented the state-of-the-art in most applications requi-
ring low-threshold and efficient lasers. O’Connor1 had
observed almost ten years ago that Nd : YV04 possessed
a larger stimulated -emission cross section at 1.06 M
than Nd :YAG and obtained iow-threshold laser operation
at 90 °K. Recently, detailed spectroscopic measurements
quantified O’Connor's observations by showing that the
stimulated-emlsslon cross section of A -axis Nd :YV04
at 1. 0634 p was 4. 6 times greater than that of Nd : YAG
at 1.0642 p.2
The superiority of Nd : YV04 lasers at 1.06 and 1. 34
p over Nd : YAG is demonstrated here by a comparison
of the laser characteristics of small spectroscopic sam-
ples of Nd : YV04 with a high -optical-quality Nd : YAG
laser rod. The comparisons were effected by end pump-
ing with cw and pulsed argon-ion lasers at 514. 5 nm,
with emphasis on cw performance. Development of
Nd :YV04 lasers Is hindered by the lack of large (3x30
mm) laser-grade crystals. However, difficulties in
vanadate crystal growth have been Identified and large
laser-grade crystals may soon be available.5
The equations describing the laser performance can
be derived from consideration of the laser-threshold
condition
RJH, exp2L{nwo - A) = 1 (1)
where R [ and R 2 are the corrected mirror reflectivities
(see explanation which follows), L is the length of the
laser sample, nm is the density of Nd3* ions in the upper
state of the User transition, o is the peak stimuUted-
emission cross section, A is the total losses per cm
(exclusive of transmission losses by the mirrors) which
include, for example, diffraction losses, excited -state
absorption, and scattering losses. The rod ends were
plane parallel and carefully aligned with the plane
dielectric -coated end mirrors. In laser operation, the
Fabry -Perot condition for maximum reflectivity
R'z= (r1*2 + R1/2)2[l + (rR)1/2]“2 should be a good approxi-
mation with r the Fresnei reflectivity of the crystal and
R the mirror reflectivity. 4
In cw operation the upper- level population is related
to the pump power according to
, _3£dVL
— »
hv>V
(2)
where rtm is the population per cm3 of the upper User
level, i? Is the quantum efficiency, namely, the fraction-
al number of Nd3* In the 4FJ/2 level per absorbed pho-
ton, fB is the fractional popuUtion in the upper User
sublevel of the 4F3/2 state, Pm Is the pump power absorb-
ed by the laser crystal, r is the fluorescence lifetime
of the 4F1/2 level, h vp is the energy per pump photon,
and V is the volume pumped. Substitution of Eq. (2) Into
Eq. (1) yields
q/j-P.T 2L<J = 2L A — InR'R' (3)
hvpV
The advantages of our method, namely, the accessibility
of the quantities in Eq. (3) to direct measurement have
been described earlier. 8 A critical evaluation of the mea-
surement techniques will be treated in a subsequent
publication.
One of the more difficult quantities to measure in a
four-ievei User system such as Nd : YV04 is the loss
factor A . This can be determined by measurement of
FIG. 1. Experimental arrangement.
Journal of Applied Physics, Vol. 47, No. 1, January 1976
Copyright © 1976 American Institute of Physics
232
232
TABLE I. Partial list of the parameters of Nd :YV04 (1%) and
Nd :YAG (l£) crystals used in the paper. Entries above the
broken line are taken from Ref. 2. The other values were
obtained in this study.
Nd : YV04
04 axis)
Nd : YAG
y (nn>)
1.0634
1. 0643
T (ms)
92
240
Ay (cm-1)
7
6
a (cm'2)
•30 x l(H*
6.5x1 0"tJ
L (cm)
0.48
1.275
V (cm3)
1.27x10-*
3.86X10*5
A (cm*1)
0.16
0.018
the laser threshold as a function of the reflectivity of the
output mirrors. 5
The experimental arrangement utilized for these mea-
surements is shown in Fig. 1. By chopping the cw argon-
ion laser beam, pump pulses of several millisecond du-
ration were provided. Since this is long compared to the
cavity and crystal response times, the results are direct-
ly applicable to cw performance of the lasers. Chopping
reduces the heat load in the crystal and also prevents
damage to the dielectric coatings on the mirrors. The
Nd :YAG rod was anti reflection coated at 1.06 ji, while
the Nd :YV04 sample was uncoated. The resonator con-
sisted of two plane -parallel mirrors; one a high (99%)
reflector and the other a partial reflector of 99, 95.6,
84, 72. 6, 67. 6, and 53% reflectivity. A plot of the
absorbed power vs - lnFt[R't was accurately linear.
The intercepts of the ordinate axis at PM = 0 provided the
loss coefficients, A = 0. 16 cm"1 for Nd : YV04 (1% con-
centration) and 0. 018 cm"1 for Nd : YAG (1% concentra-
tion). The large losses for Nd : YV04 were attributed to
scattering by the iridium platelets in the crystal which
were introduced during the growth process.3 A partial
list of the parameters of Nd : YV04 (1%) and Nd : YAG
(1%) crystals used in this paper is given in Table I.
FIG. 2. Output power of Nd :YV04 (1%) and output power of
Nd :YAG vs absorbed 0. 5145-p Input power.
FIG. 3. Output power of Nd :YV04 (1%), Nd :YV04 (2%), and
Nd :YAO (1%) vs absorbed 0. 5145-p Input power.
Comparison of the laser performance of Nd:YV04
and Nd : YAG with optimum output mirrors of reflectivity
of 67.6 and 84%, respectively, is shown in Fig. 2. The
dashed line in Fig. 2 is obtained by assuming a loss co-
efficient for the Nd :YV04 sample (heavy solid line) equal
to that of the Nd : YAG rod and is illustrative of the su-
perior performance of Nd : YV04 to be expected with
low-loss material. The slope efficiency of Nd : YV04 at
1.06 p (Fig. 2) for optimum coupling was 20%. Some
advantages of investigation of laser properties by laser
end pumping are evident by the well-controlled perfor-
mance and the exact linearity of the graphs of Figs. 2
and 3. When the laser is operated close to threshold,
the output power of the laser is given by*
P = P9[(W/Wt1) - l], where P is output power, Pa is a
constant which depends upon the material parameters,
W is the pump power absorbed, and Wih is the pump
power absorbed at threshold. A difficulty was encoun-
tered in that optical feedback produced by reflection of
pump light back into the argon-ion laser made optimum
adjustment of the apparatus critical. A Faraday rotator
is currently under construction which will isolate the
argon-ion laser from the test laser and should circum-
vent this difficulty in future experiments.
There has been considerable interest in the develop-
ment of eye-safe lasers, and this consideration prompted
our investigation of the performance of Nd : YV04 at 1. 34
M. Continuous -wave operation of Nd : YV04 at 1. 34 p was
readily achieved. A comparison of the cw operation of
Nd :YV04 (1 and 2% samples) and Nd : YAG (1%) is shown
in Fig. 3. The samples used in Fig. 3 were the same as
in Fig. 1, except for the 2% Nd : YV04 sample.
The experimental arrangement was similar to that of
Fig. 1, except that a 10 W argon -ion laser was used to
provide pump power at 0. 514 ji. The resonator consist-
ed of a pair of plane-parallel mirrors both of 99% re-
flectivity at 1.34 p which were specially designed for
low reflectivity (less than 10%) at 1.06 ji, thereby sup-
pressing oscillation at this wavelength. The superiority
of the Nd : YV04 samples is evident despite .their greater
233
J. Appl. Phys., Vol. 47. No. 1, January 1976
Tucker et 9f.
233
losses. The slope efficiency of the 2 % Nd : YV04 at 1. 34
p was 7%.
Work is in progress to determine the loss coefficient
at 1.34 p and to measure the stimulated-emission cross
sections of Nd : YV04 and Nd : YAG at 1 . 34 p. A prelimi-
nary estimate of the stimulated-emission cross section
of Nd : YV04 (1%) at 1. 34 p (assuming a A of 0. 16 cm'1)
and using the data of Fig. 3, shows that it is consider-
ably Larger than that of NdrYAG.
In adjustment of the mirrors of the Nd : YV04 lasers,
in the quasi-cw mode of operation at 1.06 p, pulsed
outputs, in several instances, were observed. The YV04
sample was misaligned with respect to the resonator
axis by about 25 arc min and self-giant pulsing at 1. 06
p was readily obtained. Pulsewidths of 40 ns were ob-
served with peak powers over 1000 times greater than
the cw output of the laser. Self- giant pulsing was ob-
served at 1.34 p, but only preliminary observations
were obtained. Self-Q-switched operation of NdrYAG
couid not be achieved under the same conditions, but
had been observed earlier by cooling the NdrYAG.7 The
ease of self-Q-switching the Nd:YV04 evidently stems
from the higher stimulated-emission cross sections.
^.R. O’Connor, Appl. Phys. Lett. 9, 407 (1906).
*M. Baas, L.G. DeShazer, and U. Ranon, Report No. ECOM-
74-0104-1, 1974 (unpublished).
*A.M. Chase (private communication).
4D. Findlay andR.A. Clay, Phys. Lett. 20, 277 (1966).
*M. Blmbaum and J.A. Gelbwacha, J. Appl. Phya. 43, 2335
a 972).
*G. Blrnbauro, Optical Mas* r* (Academic, New York, 1964),
p. 90.
tM. Btrnbaum and C.L. Fincher, Proc. IEEE 57, 804 (1969).
234
J. Appl. Phys., Vol 47. No. 1, January 1976
Tucker et at.
234
DISTRIBUTION LIST
101 Defense Documentation Center
ATTN: DDC-TCA
Cameron Station (Bldg 5)
012 Alexandria, VA 22314
107 Director
National Security Agency
ATTN: TDL
001 Fort George G. Meade, MD 20755
108 Director, Defense Nuclear Agency
ATTN: Technical Library
001 Washington, DC 20305
200 Office of Naval Research
Code 427
001 Arlington, VA 22217
201 Cdr, Naval Ship Systems Command
Technical Library, Rm 3 S-08
National Center No. 3
001 Washington, DC 20360
205 Director
Naval Research Laboratory
ATTN: Code 2627
001 Washington, DC 20375
206 Commander
Naval Electronics Lab Center
ATTN: Library
001 San Diego, CA 92152
207 Commander
US Naval Ordnance Laboratory
ATTN: Technical Library
White Oak, Silver Spring
001 MD 20910
210 Commandant, Marine Corps
HQ, US Marine Corps
ATTN: Code LMC
001 Washington, DC 20380
211 HQ, US Marine Crops
ATTN: Code INTS
001 Washington, DC 20380
212 Cmd, Control & Communications
Div Development Center
Marine Corps Dev & Educ Cmd
001 Quantico, VA 22134
Naval Telecommunications Command
Tech Library, Code 422
4401 Massachusetts Avenue, NW
Washington, DC 20390
Naval Research Laboratory
Code 6503, LTPO
Washington, DC 20390
Rome Air Development Center
ATTN: Documents Library (TILD)
Griffiss A FB, NY 13441
HQ ESD(XRRI)
L. G. Hanscom Field
Bedford, MA 01730
A FSPCOMMCEN/SUR
San Antonio, TX 78243
Armament Development & Test Ctr.
ATTN: DLOSL, Tech Library
Eglin Air Force Base, FL 32 542
HQ, Air Force Systems Command
ATTN: DLCA
Andrews Air Force Base
Washington, DC 20331
Air Force Systems Command
Advanced Systems Div (ASD/ENAD)
Wright-Patter son A FB, OH 45433
Commander
HQ, TRADOC
ATTN: A TTNG-XO
Ft Monroe, VA 23651
HQDA (DACE-CMS)
Washington, DC 20310
OSASS -RD
Washington, DC 20310
Commander
US Army Training & Doctrine Cmd
ATTN: ATCD-SI
Fort Monroe, VA 23651
215
001
220
001
301
001
307
001
312
001
313
001
314
001
320
001
401
001
403
001
405
001
406
001
(1)
407
001
408
001
409
001
410
001
411
001
413
001
416
004
419
001
420
001
421
002
423
001
Commander
US Army Training & Doctrine Cmd
ATTN: ATCD-CI
Fort Monroe, VA 23651
HQDA (DARD-ARS-P/
Dr. R.B. Watson)
Washington, DC 20310
Cdr, US Army Materiel Command
ATTN: AMCMA-EE
5001 Eisenhower Ave
Alexandria, VA 22333
Cdr, US Army Materiel Command
ATTN: AMCRD-TP (Dr. B. Zarwyn)
5001 Eisenhower Ave
Alexandria, VA 22333
Cdr, US Army Materiel Command
ATTN: AMCRD-W
5001 Eisenhower Ave
Alexandria, VA 22333
Cdr, US Army Materiel Command
ATTN: AMCRD-FW
5001 Eisenhower Ave
Alexandria, VA 22333
Commander
US Army R&D Group (Far East)
APO, San Francisco 96343
Commander
US Army Missile Command
ATTN: AMSMI-RRA, Bldg 7770
Redstone A rsenal, AL 35809
Commander
US Army Missile Command
ATTN: AMSMI-RR
(Mr. W. B. Jennings)
Redstone Arsenal, AL 35809
Cdr, US Army Missile Command
Redstone Scientific Info Center
ATTN: Chief, Document Section
Redstone A rsenal, AL 35809
Commander
US Army Armament Command
ATTN: AMSAR-RDP (Library)
Rock Island, IL 61201
Commander
US Army Aeromedical Research Lab
ATTN: Library
Fort Rucker, AL 36362
Commander
Rock Island Arsenal
ATTN: SARRI-LP-L-Tech Library
Rock Island, IL 61201
Commander, Rock Island A rsenal
US Army Armament Command
ATTN: SWERI-RET-E
(Dr. L. Gardner)
Rock Island Arsenal, IL 61201
Cdr, US Army Combined Arms
Combat Developments Activity
ATTN: A TCAIC-IE
Fort Leavenworth, KS 66027
Commander
US Army Logistics Center
ATTN: ATCL-MA
Fort Lee, VA 23801
Commander
US Army Intelligence School
ATTN: ATSIT-CTD
Fort Huachuca, AZ 85613
Cdr, US Army Aviation Systems Cmd
ATTN: AMSAV-G
PO Box 209
St. Louis, MO 63166
Commander
US Army Test & Evaluation Command
ATTN: A MSTE-D5-E (Mr. J. Bialo)
Aberdeen Proving Ground, MD 21005
Commandant
US Army Engineer School
ATTN: A TSE-CDT-DT-TL
Fort Belvoir, VA 22060
Commander
Harry Diamond Laboratories
ATTN: A MX DO-RCB, Mr. F. Harris
Washington, DC 20438
422
424
001
425
001
427
003
429
001
431
001
433
001
434
001
436
001
441
001
(2)
442 Commander
Harry Diamond Laboratories
ATTN: Library
2800 Powder Mill Road
001 Adelphi, MD 20783
447 Cdr, US Army Picatinny Arsenal
ATTN: SMUPA-VC5
Bldg 350
001 Dover, NJ 07801
449 Commander
Picatinny Arsenal
ATTN: SARPA-TS-S #59
002 Dover, NJ 07801
451 Commander
Frankford Arsenal
ATTN: Dr. Wm McNeill, PDS64-4
001 Philadelphia, PA 19137
454 Cdr, US Army Training Device Agcy
ATTN: AMXPG
Naval Training Equipment Center
001 Orlando, FL 32813
456 Commander
White Sands Missile Range
ATTN: STEWS- R E - IO
001 White Sands Missile Range, NM 88002
458 Dir/Dev & Engr
Defense Systems Div
ATTN: SAREA -DE-DDR,
H. Tannenbaum
002 Edgewood A r senal, APG, MD 21010
465 Commander
Aberdeen Proving Ground
ATTN: STEAP-TL (Bldg 305)
002 Aberdeen Proving Ground, MD 21005
480 Commander
USASA Test & Evaluation Center
001 Fort Huachuca, AZ 85613
483 Cdr, US Army Research Ofc
ATTN: AMXRO-IP
PO Box 12211
001 Research Triangle Park, NC 27709
484 Cdr, US Army Research Ofc
ATTN: AMXRO-PH (Dr. R. J. Lontz)
PO Box 1221 1
001 Research Triangle Park, NC 27709
485 Cdr, HQ MASSTER
Technical Information Center
ATTN: Mrs. Ruth Reynolds
001 Fort Hood, TX 76544
486 Commander
US Army Mobility Eqpt R&D Cen
ATTN: SMEFB-R
001 Fort Belvoir, VA 22060
488 US Army Security Agency
ATTN: LARD
Arlington Hall Station
001 Arlington, VA 22212
489 Commander
US Army Tank- A utomotive Command
ATTN: AMSTA -RW-L
001 Warren, MI 48090
492 Commandant
US Army Air Defense School
ATTN: C&S Dept, Msl Sci Div
001 Fort Bliss, TX 79916
496 Cdr, US Army Combined Arms
Combat Developments Activity
ATTN: ATCACC
001 Fort Leavenworth, KS 66027
500 Commander
US Army Yuma Proving Ground
ATTN: STEYP-MTD (Tech Library)
002 Yuma, AZ 85364
501 Commander
US Army Arctic Test Center
ATTN: STEAC-PL
002 APO, Seattle 98733
502 CO, US Army Tropic Test Center
ATTN: STETC-MO-A (Tech Library)
Drawer 942
001 Fort Clayton, Canal Zone 09827
(3)
511
001
512
001
514
001
517
001
518
001
519
001
520
001
596
002
603
001
610
001
Commander
US Army Nuclear Agency
Fort Bliss, TX 79916
Directorate of Combat Dev
US Army Armor School
ATTN: ATSB-CD-AA
Fort Knox, KY 40121
Commandant
US Army Inst for Military
Assistance
ATTN: A TSU-CTD-OMS
Fort Bragg, NC 28307
Commander
US Army Missile Command
ATTN: AMSMI-RE (Mr. Pittman)
Redstone Arsenal, AL 35809
Commander
Harry Diamond Laboratories
ATTN: AMXDO-RCB
(Dr. J. Nemarich)
Washington, DC 20438
Commander
US Army Systems Analysis Agency
ATTN: AMXSY-T (Mr. A. Reid)
Aberdeen Proving Ground, MD
21005
Cdr, Frankford Arsenal
ATTN: SMUFA -N-41 00
Bldg 201
Philadelphia, PA 19137
Commandant
US Army Signal School
ATTN: A TSN-CTD-MS
Fort Gordon, GA 30905
ECOM Liaison Office
US Army Electronic Proving Ground
Fort Huachuca, AZ 85613
Director, Night Vision Laboratory
US Army Electronics Command
ATTN: AMSEL-NV-D
Fort Belvoir, VA 22060
607 Commander
US Army Tank -A utomotive Cmd
ATTN: AMSTA-RHP, Dr. J. Parks
001 Warren, MI 48090
608 Director
Night Vision Laboratory (ECOM)
ATTN: AMSEL-NV-VIS
(Mr. L. Gillespie)
001 Fort Belvoir, VA 22060
609 Director
Night Vision Laboratory (ECOM)
ATTN; AMSEL-NV-SD (Mr. Gibson)
001 Fort Belvoir, VA 22060
614 Chief
Ofc of Missile Electronic Warfare
Electronic Warfare Lab, ECOM
001 White Sands Missile Range
NM 88002
617 Chief
Intel Materiel Dev & Support Ofc
Electronic Warfare Lab, ECOM
001 Fort Meade, MD 20755
680 Commander
US Army Electronics Command
000 Fort Monmouth, NJ 07703
1 AMSEL- PP-I-PI
1 AMSEL- PL-ST
1 AMSEL- NL-D
1 AMSEL- WL-D
1 AMSEL- VL-D
1 AMSEL-BL-D
1 AMSEL-CT-L (Dr. R. Buser)
1 AMSEL-CT-LC (Dr. E. Tebo)
1 AMSEL-CT-L (Mr. B. Louis)
1 AMSEL-CT-LD (Dr. E. Schiel)
1 AMSEL-CT-LE
(Dr. H. Hieslmair)
4 AMSEL-CT-LD (Mr. J. Strozyk)
1 AMSEL-TL-D
1 AMSEL-TL-I (Dr. H. Jacobs)
1 AMSEL-TL-B (Mr. M. Zinn)
1 AMSEL- TL-TG
(Mr. S. Schneider)
1 AMSEL-NL-R -5
(Dr. E. Dworkin)
(4)
701
001
703
705
002
706
002
707
001
708
001
711
001
718
001
720
001
001
ATTN: Acquisitions Br (S-AK/DL)
PO Box 33
College Park, MD 20740
Advisory Group on Electronic Devices
201 Varick Street, 9th Floor
New York, NY 10014
Advisory Group on Electronic Devices
ATTN: Secy, Working Group D
(Lasers)
201 Varick Street
New York, NY 10014
Target Signature Analysis Center
Env. Research Inst of Michigan
PO Box 618
Ann Arbor, MI 48107
Ballistic Missile Radiation Anal Cen
Env. Research Inst of Michigan
Box 618
Ann Arbor, MI 48107
Metals and Ceramics Inf Center
Battelle
505 King Avenue
Columbus, OH 43201
1 AMSEL-VL-E
1 AMSEL-WL-N
1 AMSEL-CT-L-D (Ofc of Record)
1 AMSEL-MA-MP
2 AMSEL-MS-TI
1 AMSEL-GG-TD
2 AMSEL-PA
1 A MCPM- TDS-SE
1 USMC-LNO
1 AMSEL-RD
1 TRADOC-LNO
1 AMSEL-CG (Mr. Doxey)
25 Originating Office
MIT- Lincoln Laboratory
ATTN: Library - Rm A-082
PO Box 73
Lexington, MA 02173
NASA Scientific & Technical
Information Facility
TACTEC
Battelle Memorial Institute
505 King Avenue
Columbus, OH 43201
Thermophysical Properties
Res Center
Purdue University,
Research Park
2595 Yeager Road
Lafayette, IN 47906
Airtron Division,
Litton Industries
Dr. R. Belt
200 East Hanover Ave.
Morris Plains, NJ 07950
(5)
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11